University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange

Doctoral Dissertations Graduate School

8-2005

Using Projective Measures to Examine the Relationship Between Adult Attachment Status and Object Relations

Betty Marie Martin University of Tennessee, Knoxville

Follow this and additional works at: https://trace.tennessee.edu/utk_graddiss

Part of the Psychology Commons

Recommended Citation Martin, Betty Marie, "Using Projective Measures to Examine the Relationship Between Adult Attachment Status and Object Relations. " PhD diss., University of Tennessee, 2005. https://trace.tennessee.edu/utk_graddiss/4367

This Dissertation is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council:

I am submitting herewith a dissertation written by Betty Marie Martin entitled "Using Projective Measures to Examine the Relationship Between Adult Attachment Status and Object Relations." I have examined the final electronic copy of this dissertation for form and content and recommend that it be accepted in partial fulfillment of the equirr ements for the degree of Doctor of Philosophy, with a major in Psychology.

Leonard Handler, Major Professor

We have read this dissertation and recommend its acceptance:

Robert Wahler, Laurence James, Richard Saudargas

Accepted for the Council:

Carolyn R. Hodges

Vice Provost and Dean of the Graduate School

(Original signatures are on file with official studentecor r ds.) To the Graduate Council:

I am submitting herewith a dissertation written by Betty Marie Martin entitled "Using Projective Measures to Examine the Relationship Between Adult Attachment Status and Object Relations." I have examined the final paper copy of this dissertation forform and content and recommend that it be accepted in partial fulfillmentof the requirements for the degree of Doctor of Philosophy, with a major in Psychology.

�Handler, MajJr Professor

We have read this dissertation recommend its acceptance:

/

Acceptance for the Council:

Vice Chancello Graduate Studie 7/(8:Jt . , JJ ., � Using Projective Measures to Examine the Relationship Between Adult Attachment Status and Object Relations

A Dissertation Presented for the Doctor of Philosophy Degree The University of Tennessee, Knoxville

Betty Marie Martin August 2005 Dedication

This dissertation is dedicated to Drs. Chuck Jones, Jack Barlow, Joyce Cartor, and Mary MargaretKelly. You have all become significant parts of my internal world through acceptance, support, opening doors, showing me exciting paths, andguiding me to new understandings. Thankyou.

11 Acknowledgments

I would like to thankDrs. Leonard Handler, Robert Wahler, Richard Saudargas, and Laurence James for servingon my doctoral committee. I appreciate Dr. Wahler's introduction to attachment theory, academic support,and continuous effortsin helping me become familiarwith measures of attachment. I would like to thankDr. Handler for his mentorship, guidance in research, and introducing me to CarolGeorge and Malcolm West. I also thankCarol George and Malcolm West for their excellent training in the use of the Adult Attachment Projective. Further acknowledgment goes to Paul Hoffman, Eric Peters, Ian Haag, Chad Sims, Noah Roost, and Yakini Stephens. This dissertation would not have been possible without their research contributions.

JI) Abstract

The purpose of this study was to document an empirical link between attachment theory and object relations by using projective measures, while furtherdefining John

Bowlby's concept of the internalworking model. The internalworking model is a set of unconscious cognitive and emotional guidelines forhow an individual understands interpersonalinteractions, and influencesbehavioral and emotional responses (Bowlby,

1973, 1988). The internalworking model, as described in Bowlby's attachment theory; bears striking resemblance to object relations theories of internalized unconscious representations. For example, Donald Winnicott stres�ed th�_importance of early interactions with the primary caregiver as shaping the child's internalmental an� emotional world through processes of internalizingrepresentations of the parental figure as a psychological object (Winnicott, 1931). 100 Undergraduates were administered the

Adult Attachment Projective (George, West, and Pettem, 1999) and Thematic

Apperception Test, measured by Drew Westen's Social Cognitionand Object Relations

Scale (Westen, 1995). Results show that individuals classifiedas secure have significantlyhigher average scores on the object relations dimensions of Complexity of

Representation of People, AffectiveQuality of Experiences, Emotional Investment in

Relationships, Understanding of Social Causality, Experience and Management of

Aggressive Impulses, Self-Esteem, and Identity and Coherence of Self than insecure individuals. These findingssuggest that object relations can be used to provide a more comprehensive understanding of the internal working model underlying adult attachment status.

IV Table of Contents Chapter Page

I. Introduction ...... I Attachment Theory...... 4 Classical PsychoanalyticTheory ...... 7 A Comparison of AttachmentTheory and Classical Psychoanalytic Theory...... 11 Object Relations Theory ...... , ...... 16 A Comparisonof Attachment Theoryand Winnicott's Object Relations Theory ...... _ ...... 19 Additional Object Relations Theories...... 22 Research on Children's Attachment...... 30 Research on Adult Attachment Status and Object Relations ...... 34 II. Method ...... 44 Participants...... 44 Materials ...... 44 III. Results ...... 50 N. Conclusion ...... 52 List of References ...... � ...... 67 Appendix A ...... -...... 76 Vita ...... :...... 86

V List of Tables

Table Page

1. Kappa and Intra-class CorrelationCoefficients Jor the Interrater Reliability of the AAP and SCORS ...... 77 2. Levene's Test of Equality of ErrorVanances ...... � .. � .... :...... \ ...... 78 3. Sample Demographic Data ...... : ...... 79 4. Pearson's R Correlations of Demographic Data and SCORS .;� .....80 5. AAP & Other Death Cross Tabulations anc;lChi-Square Test...... 81 6. AAP& College Year Cross Tabulattons and Chi-Square.Test .-. ....82 · 7. AAP & Parental Divorce Cross Tabulation� and Chi-Square Test.83 8. Descriptive Statistics and Estimates of A.APand SCORS ..·u.-: •••••••84 9. Comparisons of Secure and Insecure Findings of the A.APand SCORS ...... ,....•..... :.�...... ; .. : ...... � .. �.�.:.·...... 85

Vt I. Introduction

In recent years, developmental psychology has beganto use attachment theory-­ specificallyan individual's attachment status--to help explain personalityand behavioral development, developmental psychopathology, and to deriveclinical implications for therapeutic intervention. Attachment theory has its origins in evolutionary, ethological, psychoanalyticand cognitive approaches to human development (Bowlby, 1988). These approaches emphasize early caregiving experiences as forcesthat shape an individual's internal working model of attachmentrelationships (Bowlby, 1988).

According to attachment theory, the initial interactions between primary caregiver and child will influencehow well attached the child is to the primary caregiver. The child will instinctually seek to engage the mother as a means to survival. As the child becomes either securely or insecurely attached to the caregiver,the child will develop internalcognitive structures based on the caregiver relationship, known as interna1 working models. These internal models, largelyunconscious, set the structure forhow the child perceives self and others in the· context of interpersonal relationships and carry implications forlater personality and behavioral development. This approach, heavily influencedby classical psychoanalytic and object relations theory ( e.g., the internalizationof interpersonalexperiences), has gained prominence forstressing the importanceof the primarycaregiver-child bond as the foundation of personality development (Sable, 1992).

Classical psychoanalytic theory and object relations theory differslightly from attachment theory in the drives which are said to motivate infantand child behavior.

1 Classical psychoanalytic theory, originally developed by SigmundFreud, emphasizes sexual and aggressive urges as instinctually motivating infantand child behavior. Later, ·. . ·. . ', .. object relations theories developed out of Freud's original theories and em h si e human· · · � � �. relatedness rather than sexual and aggressive drives. According to object rela�ons · · · · · theories, the child has an inherent desire to engage in re�ationships with objects in the world, such as humans. While object relations places less emphasis on instinctualdrives, · both classical psychoanalytic and object relations theories argue that the _early relationship with the primary caregiver leads to mental andemotional internalizations of self and ·relationship-oriented representations which later influence psychological development. Bowlby originally arguedthat his work was rooted in classical psychoanalytic andobject re1ationa1 theories, but these theories suffered fromJack of empirical research and scientificmeasures to tap unconscious-processes (1969, 1988).

While attachment theory has generated a rich anq rapidly growing body of ; findingsaddressing empirica] understandings of infan�_attachments and related childhood personality styles and behaviors, there has been little work applying researchmethods to scientifically document the relationship among attachment theory, classical psychoanalytic theory, and object relational processes. Such research is inherently importantto the fieldof psychology because classical psychoanalytic theory and object relations theories have yet, forthe most part, to earn scientific acceptance in the field of behavioral sciences due to a lack of empirical support forthe theories. Both cl�sical psychoanalytic and object relations theories richly inform clinicians and therapists in the psychological processes which underlie personality and behaviora] development, but the

2 :.·' ' , . validity of such theories remains in question when there is a significantlack of scientific

. ' • ... • .� ·\ ' . ,• •! ' . . ·.. evidence to document complex, dynamic, and oftenunconscious psychological phenomena.

Thoughattachment theory has significantlycontributed to the fieldof psychology through expandingunderstandings of infantand child psychological functions, comparatively less research has been generated in the realms of adult attachm�nt and personality functioning (Rothbard & Shaver, 1994). For example, although there has been discussion and research of adult attachment, there are fewstudies which ]ink adult attachment statusto daily psychological functioning(Rothbard & Shaver, 1994). In this arena, object relations researchhas expanded beyond attachment theory to use empirically derived measures to study object relations in the differentiationof psychopathology.

Rather than continue segregation in paradigmsof thinking, th.is study proposes that integrationaf the major theories would allow fora more comprehensive understanding of human development, with implications foradult personality functioning and psychopathology. Focusing on the strengthsand commonalities among the theories would generate directions for future research, increase knowledgeof psychological growthprocesses, and influence therapeutic interventions forpsychopathology. In this way, attachment theory, classical psychoanalytic theory,and object relations theory can be interpretedto informand expandupon each other to promote increased awareness of ... : . psychological processes.

3 Attachment Theory

Attachment theory, as developed by John Bowlby, states that infants are bornwith instinctual motives to demonstrate goal-oriented behavior in a particularenvir onment, namely to engage mother as a means to survival(Bowlby, 1969). Such behaviors are affect laden while the infant interprets and processes sensory information from his/her internalworld and externalenvironment to ensure survival(Bowlby, 1969). For example, the baby cries when hungryto alert mother so the baby can be fed(Bo wlby,

1969). Bowlby theorized that such behaviors on the partof the infant then solidify the relationship with the primary caregiver to meet the physical and emotional needs of the child, protect the child fromharm, and educate the child in gradual autonomy (Bowlby,

1969). He believed that the child was particularlyvulnerable to the effects of either adequate or inadequate care during the first several years of the child's life(Bowlby,

1969). Acr..ordingto Bowlby, a primary caregiver who is attuned to the infantand meets the infant'sphysical and emotional needs in a timely f�hion will enable the infantto

·cognitivelydevelop a functioninginternal working model to understand and predict attachment oriented behaviors in the self and others (Bowlby, 1988).

The internalworking model is an internal, cognitive, often unconscious mental structure, which sets guidelines forhow the infantunderstands self and others in the context of interpersonal interactions (Bowlby, 1988). The working model will then influencethe infantand developing child in perceiving events, forecastingthe future, and constructing plans or behaviors and largelyinfluence emotionalresponses (Bowlby,

1973). This strong emotional bond between caregiver and infantwill eventuallyenable

4 the infantto internalizecomforting responses fromthe caregiverin times of distress so

that the infantcomes to learnemotional self-regulation skills.(Bowlby, 1988). ·For example, because the baby experienced attuned caregivingfrom her mother, she can

come to rely on her mother in times of need and expect to be soothedby her mother.

This level of stability and predictability enables the infant to internalizea concept of her·· mother which is comforting, eventually leading the infantto call upon thisrepresentation

in times of distress without needing the actual physical presence of her mother. She riow

has an internalworking model of mother, which may facilitatethe development of self­ soothingbehaviors such as thumb sucking or holding a stuffedtoy to comfortherself.

A major disruptionin primary caregiving� such as loss of or change in a caregiver,

or the experience of a caregiverwho is unpredictable, neglectful,rejecting or abusive will inhibitthe development of a functioningand adaptive internalworking model in the

child, while facilitatingthe increased likelihood of developmental ·psychopathology (Bowlby, 1988). Bowlby emphasized the role of emotion in attachment theory as the. ·

child may become overwhelmed with anxiety or mourningin light of disrupted or poorly·

attuned caregiving ( 1988). He postulated that if a child is physiologically overstimulated due to disruptionsor inadequacies in the caregiving experience, the child will attemptto

adapt to and contain the overstimulation by altering the -flow of sensory information,

creating long lasting implications forpersonality development {Bowlby, 1980). Such·· ·

adaptations can include ignoringor distorting cognitive and perceptual· information

through repression, cognitivedisconnection froma situation, redirection of information,

or preoccupation with personal sufferings(Bowlby, 1980). Repression would involve an

5 inability to recall certainthoughts and feelings·fr om memory; cognitive disconnection is

. . an inability to trace certain thoughts and feelingsto an original situation; redirection often . involves directing anger away fromthe person who elicited the emotion; andfocusing ori . , . personal sufferingis intended to direct attention away. from � more difficultand painful

: ' situation (Bowlby, 1980). These distortions thenimpact and distort th� struchir�of the internalworking model of relationships and create child attachment relationships which . maybe insecure, anxious, over dependent, and angryrather 'than se·cure (Bowlby� .1973).··

Oncethe child has developed maladaptive internal�orking· m6d�ls b-�secion the original caregiving experience, his/her ability to function in.futurerel ationships� or even· act effectively in the world, is compromised. The child, forexample; now has an· internal . working mode) that is not rooted in stability, predictability, or the expectiltiortthat the caregiver can soothe. Rather, the internal working model �nay be wourtdediii ' abilse, . neglect, loss, fear,anxiety, or anger (Bowlby, 1973). The m·aladaptive Internal�orking _ · model will then influencehow the child perceives self, understimds'others:in the·context . ··:

. . -;,, . of the world, and colors most interpersonalinteractions (Bowlby, 1973). Pathology can then involve a compromised understanding of the self arid·oth�rs, · 1eading. to· a host of . psychological inadequacies, complications, and disorders (Bowlby, .1973). Psychological disturbanceis related to disruption of main attachme�t relationships. or 'an' inability to formoriginal attachment relations,while treatment interventi�ns· 'address' the .impac( of attachment; current and/or past separation and loss experie�ces, with. the' inte�i to' modify' thinking, feeling, and behavior (Bowlby, 1988). Bowlby's att�chm��t theoryhas · . significantlyimpacted the fieldof developmental psycho��gy through the pr��ision of a :

6 frameworkof understandingand researchinginfant attachment, consequent childhood

personality styles and behaviors, and continued emphasis on· earlychildhood experiences·

as fundamental to personality development. Meanwhile, Bowlby stressed that the infant

from birth has an arrayof potentiaUy available developmental pathwaysor cours·es of · ·

human development (Bowlby, 1988). Progression along a particularpathway is·

determined b¥ environmental factorsand the way in which child contends with those

factors(Bowlby, 1988). A child with hea1thy parental relations will most likely develop

healthy internalworking models and consequently have improved psychological .health

than a child with poor parental relations and maladaptive internalworking models. (Bowlby, 1988). Even so, there exist continuing potentials forchanges in growth-and

developmentin either positive or negative directions for most any individual throughout

.the lifecycle (Bowlby., 1988).

Classical Psychoanalytic Theory

Sigmund Freud has informed the fieldofpsychology through his various writings

between 1888-1937. He was the firstpsychologist -who introduced the menta] and

emotional world of the infant as well as the importance of the early relationship between

infant and primary caregiver. Freud also significantlycontributed to the fieldof

psychology the roJe of traumain personality development and the implications of ·

unconscious pro�esses. Further, he provided major insights into the development and

treatment of psychopathology and laid the groundwork forlater theories in object

relations.

7 Freud originally arguedthat an infantwas bornwith a �ertain �mountof psychological and physical energy, which would serv.e.to motivate th� chil� to meet . certain needs such as hunger, breathing, and sexuality �reud, 189�). For_tlle_mo�� p�,

.e the energy would remain in a state of inertia, until sue� neecls cre�te_d a dist��at\�� i� tp . . child, forcingthe child to manifestbehaviors to meet those needs (Freud, ,1895). Freud . . .· . . ·. believed the infant was born with a "pleasµre principle," �lso known _late� as "libido .or libidinal energy," a psychological drive that moves the infanttoward� p_leasurable experiences rooted in sex and aggression, such as sucking the _mother's.breast �pile causing the infant to draw away from sensationsthat are n�t pleas_ur�ble (Freud, 191 l; .. Freud, 1924). Sucking the mother's breast is both sex��l �nd aggressivein nature _ according to Freud (1924). Because the infant's need� 01ustbe met frome,xtem �l . sources, a "reality principle" develops wllichaids _the �nfant in perceiving �nd m���lating. the external world, namely the relationship with the mother (F:reud, 191 1 )� Ther eality principle, later known as "ego," would also serve to keep libidinal. energie� at bay . , ...... because the infant can only tolerate a_ certain level of excitement in a given pe�o_d of . time, and can only expend a certain amount of energy at ��y point in time (Freud, 1920).

The ego is grounded in instincts geared towards self preservation a_nd will.serve to modulate the infant's internalworld and external reality to decrease tension cr�at�d �y . the conflictof wants,desires, and passions and not pavi11g �hen1 dire�tly gratifie� (Freud,

1920).

Freud also believed the amount of tension an. infant. experience_d w�ql� strongly , augment how the libido and ego function. Should an infant exp.erience too much tension .. I I •� .- ' • • • . ' • ' '• • • ' • • •

8 due to over excitation fromthe externalworld, such as the mother either over gratifying� or neglecting the child's needs, the flowof energy will be altered in an attemptto compensate forthe excessive level-of excitement (Freud, 1920), Theinf antwill employ defensive mechanisms,such as repression or forgettingunwanted andpainful ·m emories in an attempt to restore the balance of energies (Freud, 1920). Theego then serves to repress painfulmemories in anattempt to protect the baby fromwhat Freud considered trauma (Freud, 1920). He argued that trauma was any level of excitation which caused a · disturbance in the flowof energies .and required the infant to employ defensive mechanisms as a means to control the floodingof over stimulation (Freud, 1920).

According to Freud, trauma would then relate to later psychological problems because the repressed experiences will be re-enacted by the individual without conscious awareness and perhaps in an unrecognizedform (Freud, 1920).

Freud later clarifiedthe psychological structures which serveto manage intrapsychic tension such as the id, otherwise known as libidinal energie&; the ego, the

• conscious and unconscious organizer of mental experiences; and the superego, which ,t emphasizes thehigher nature of man such as religion, morality,. and-social sense of responsibility (Freud, 1923). Freud theorized that many experiences and influences which serve to shapepersonality (i.e., the flowof energies managing the struggle among the id, ego, superego, and the externalworld) are largely unconscious and rooted in the early experiences and-interactions with mother (Freud, 1923). Through processes of identificationwith the parental figure and then internalizationof those identifications,the

dynamic interplaybetween id, ego, andsuperego become solidified before the age of five

9 (Freud, 1924). As the child identifiesor finds similarities with the· parental· figure, representations oftho�e parentalobjects become internalizedin .the child's'world through � the structures of id, ego, and superego. The typeof relationship. with "the parent,· for example, mother, would then be continuously and unconsciously experienced in the child's internal world.

Freud emphasized the relationship with mother as the "love object," where the · infantis driven to expend energies in seeking pleasurable sensations� stichas feedingor · being touched (Freud, 1924). In this sense, the infantcomes to love the·mother as_ she meets those needs, and will experience intense anxiety at feeling·a loss of her (Freud;

1924). This fear is soothed by the physical presence of the other, and.will later manifest into adult neurotic fearsif comfort is not given to thechild (Freud,- 1924). In this way� : · the typeof relationship with the mother will be expressed throughthe level of conflict' among the id, ego, and superego (Freud, 1924). An adequate relationship with the mother would result in low conflictamong the internalmental structures,'while an inadequate relationship with the mother would result in more intrapsychic ·conflict ,.,

(Freud, 1924). The mother relationship then becomes the prototypefor all other later . love relationships, as the adult tries to recapture the happiness 'once experienced with the mother or re-enacts the disturbed relationship with themother (Freud, 1924 ). Although,· ·

Freud arguedthat the mother's relationship with the infantis inherently sexual in fondling and feedingthe infantand due to the infant'sdesire for sexual gratification, he believed the infant was protected fromsexual abuse due to the cultural barriers of society

.. (Freud, 1924). . _,

10 A Comparison of AttachmentTheory and C_lassical PsychoanalyticTheory

Bowlby, himself, argued that his theories we�e he�vily influenced by the °"'.'orks of Freud and stated that his theorywas only a variant of Freud's original hypothes�s ..

(Bowlby, 1969). Bowlby argued that although both theories share many common themes . .

such as sensitive periods of the infant's early life,the emphasis on early relation��i:p_s

with primary caregiver, and the pathogenic potential of loss an� tr�uma, he diffc�e� from

Freud in a few fundamentally differentways (Bowlby,_ 1969)._ First, whereas Freu� retrospectively gathered data in the traditional method of inquiry of his c_lients, Bowlby

advanced the field of psychology by gathering data from real life situations through

prospective and direct observations of children as a means t� predict later psychological . . . \ . . �. . development (Bowlby, 1969). Bowlby felt he was much mor�_spe�ifi�regarding t�e .

definitiono( trauma as being separation from or loss of the primary car�giver, �hereas

Freud did not specifically define the nature of trauma to the infant beyond a level of -......

excitation exceedingthe infant's capacity to manage (Bowlby, 1969). 1:lowlbyagre�d ...... -

with Freud that there would be stru�tural psychological configurations, psychologica_l origins to those configurations, and an adaptive component of the configurations to the r ' ' ' ' • ' • • ' ,�, • • ..,

environment (Bowlby, 1969). He mainJy di�agreed with Freud's understand�n�� of

psychological energy. He could i:iot conceive of an infant haying a s��d_am��nt of

psychological. energy which would be expended in ce�in amounts at given periods of ' ' .. • r. ' • time (1969). Hence there is a primary and fundamental difference in what motivates . . . , . . ,·.· .·. infant and child behavior, internal psychological and sexual energies., or the instinctual desire to seek mother as a means of survival. Bowlby also rejected the concept that an

11 individual passes through stages of development, and can become fixatedat or regressto a particular stage ( 1988). He emphasized the ip-iportanceof multi-varied developmental pathways being open to a child, and that behaviors viewed as neurotic or fixatedfrom a psychoanalytic perspectiveare actually normal behavior demonstrated by frightenedor distressed children (Bowlby, 1988).

What may initially appear to be a small differenceof opinion in what motivates behavior actually created a major break between Bowlby and other psychoanalytic thinkers (Fonagy, 1999). Fonagy noted thatattachment theorywas criticized by psychoanalyticthinkers for being too mechanistic, nondynamic, renouncing drives, unconscious processes, and ignoringcomplex internalizedmotivational and conflict resolution systems (Fonagy, 1999). Attachment theoryalso seemed to ignorethe richness of human affect,as well as the infantas a plea&ure•seeking creature, with a too narrow definitionof physical separation creating vuln�rabil�ties in the child (Fonagy, 1999).

Bowlby was criticized forfailing to consi9er the role of the child's ego in navigating . t . attachment relationships and-reactions �o loss, ignoring negative attachments, and ignoring trauma other than physical separation (Fonagy, 1999). While Fonagy also discussed Bowlby's remarks on similarities between psychoanalytic and attachment theories, he also noted that Bowlby felt psychoanalytic theory gave too much weight to metaphysical workings of the mind, which could not be measured (Fonagy, 1999). In this sense, Bowlby argued that Freud placed too much emphasis on the role of fantasy in the child's life, rather than real liteexperiences such as trauma or loss (Bowlby, 1988).

12 Fonagy observed that both Freud andBowlby speak to unconscious processes, namely in Freud's psychic energies playing between id, ego, andsuperego andBowl by's internalworking models (Fonagy, 1999). Just as the ego creates defenses, such as repression, Bowlby's internalworking models will create perceptual and cognitive distortions necessary to tolerate anxiety, fear, separation, and loss (Fonagy, f999).

According to Fonagy, both Freud.andBowlby speakto biologically determined experiences such as fear, linked to the danger of losing the primary caregiver or fundamentallove object (Fonagy, 1999). He also stressed the commonalities outlined earlier by Bowlby: a) recognizingthe importance of attachment-to the mother figure, b) · the use of abandonment and isolation distressthe infant, c )-the ·idea that anxietyis tooted in fear ofloss ofthe mother, and d) the child's relation to mother is unique and laid down unalterably at an early age while becoming the prototype forlater lov� relations (Foriagy,

1999). Fonagy argued thatmain differences resided_i� the �eas of Freud's emphas� on cultural influences, the factthat his views of infancy were forthe most partabstract and not based on direct observation, his lack of differentiation ·between interactions involving real parents versus fantasy parents, his lack of clarityinvolving the development of the self, and his emphasis on repetition of childhood experiences in adulthood ( 1999).

Other theorists continue to note differences �n the t\yot1!eories. Cassidy (1998) points out that Bowlby argues instinctual drives such as_ Sex and aggression are outside· the realm of attachment theory. Sable (1992) argues that Bowlby replaced instinctual sexual andaggressive drive theorywith an ethological orientation where the infanthas instinctual patternsof behaviors that facilitatethe formingof anaff ectional bond between

13 f caregiverand infant. Thisbond allows a· continual unolding 'or'personality stiorigiy' influencedby attachment experiences, leaving the iridividuafpron"e' to health or psychological adaptation (Sable, 1992). Meanwhile, Si�verman (1991) argues that activation of the attachment system is similar to drive theory in that'experiences of the infantare internalizedas a psychological structure in the form of object represe�tations which can be distorted when interacting with drive defenses (Silverman,'199 i ).

Although F onagy argued that it is difficultto compare. classicai psychoanalytic .:_ . theory to attachment theory due to Freud's diverse and wide' reaching contributi��s to psychology, there is neverthelessroom fordisagreement. ·. In line with"Silvemian, at the heart of these two theories are fundamentally similar processes.' : Toe·intimate ··rei�tionship between mother and infantis fundamentalto personality' developm�nt because the;'. '' interactions between mother and infantlead to internalizedunderstandings of �e'tiand ···. others. Bt,wlbyand Freud both argue forthe development of psychologfoal -coristructs. . . such as internalworking models or the level of conflict between'id, ego, and superego which serve to shape future perceptions of events, interpretations·of those events, and shape behavioral and emotional responses·. Although they-may 'disagree· on 'the type of .·. . psychological drive that motivates infant behavior, the outcome is the same.· The infants needs and depends on the mother and is vulnerable· to s'evere anxiety and ·loss should primary caregivingre lationship become disturbed either throughseparation, loss: �r pathogenic care.

- With this in mind, Bowlby does not clearly elaborate on how the internal models may shape future behaviors or acknowledge their powerful affectivenature in shaping an

14 individual's reality.- Thisis where classical psychoanalytictheory significantly contributesto understandingsof personality development. Freud's model of conflict among the id, ego, and superego ar� entirely.affective in nature andexert strong influences in shaping personality and behavior to the degree that an individual can compulsively repeat earlierexperiences lived out with the parentalfigures. A parsimoniousintegration of these two theories would emphasize the following: I) the strong influenceof the early relationship between parent and child as fundamentalto human development through, 2) parent child interactions that lead to the internalization· - of the relationship in various forms, 3) these internalizationscan be disrupted, inhibited or distorted by trauma, and 4) create the need fordefensive adaptation andlater maladaptive personality and behavioral functioning. In this sense, adequateparenting would lead to healthy internalizationsof cognitiveinternal working models and affective psychological structuressuch as id, ego, superego which.are-low in conflict, enabling the child to understandpositive experiences of self and other in the world.

Meanwhile either separation fr�mor loss of the primary caregiver;or parenting .­ which was traumatic through emotional and/or physical abuse or neglect of the child, would reS;Ultin maladaptive interoalizations of the relations� Thechild experiencing such . trauma then employs defensiveprocesses as a means to adapt to the i11tolerable affective state which resulted from the failingenvironment. Such adaptations among·the internal· working models and intrapsychic structures are both cognitive and affective, leading to distorted or inaccurate understandingsof self and others, ,n.aladaptive personality

15 y functioning and behavioral functioning and the possibilit of repeating the earlier lived experience.

Object Relations Theory

Object relations theory also stresses theimportance of the early relationship with the mother as an imperative influenceon personality development and overall psychological functioning. Object relations. theories emphasize early.interactions with the primarycaregiver and determine the structure of the child's internal mental world through processes of internalizing representations of the parental figures as psychological objects. Many object relations theoristsdeviated fromFreu d�s emphasis·-on instinctual drives and child sex�Jity in order to highlight the importance of human relatedness, the desire to relate to other human_beings �- the primary motivator-ofinfantand chil4·-·· behavior. ·This paper will focuson the Independent British Object Relations work of ·_

Donald Winnicott, a pediatrician who spent ov�r twenty-five years observing- children in his practice and developed object relational theories rooted in the relationship between mother and child. An emphasis is placed on the Donald Winnicott?s theories due to marked similarities to attachment theory. The paper will also provide a -brief review of object relations theorists fromother various backgrounds.

Winnicott argued that the infantwas not capable of being an individual due to its absolute dependence upon the mother(Winnicott, 1935). The child's initial sense oflove is governed by physical needs, the desire to satisfy thoseneeds, -seeking pleasure,and . avoiding discomfort (Winnicott, 1935). The mother as object then; plays a pivotalrole:in

16 providing fo r and-satisfying the infant�s- needs while removing discomfort(Winni cott,

1935). How the mother goes about fulfillingthe needs of the infantin the firstfive years

of lifethen lays the· foundations formental health or psychoneurosis because she becomes an internalized object in the child's psychological world (Winnicott, 1931)� · The parents' attitudes, particularly the mother's, towards t�e �hi!d's inheren� greed and destructive tendencies will influence·the baby's behaviors and expectations from the world (1941 ): If the infanthas happy experiences in a stable, friendlyatmo sphere then · the infantcan build confidence in people in the external world and have belief in good things and object relations inside him/herself, which creates an overall sense of stability

(Winnicott, 1941).

Themother then becomes an internal source of love and protection for the infant, while the infant will experience fear of losing mother due to his/hernatural destructive tendencies (Winnicott, 1948). The mother becomes a presence in the child's internal world due to her ability to provide forthe child and tolerate the child's aggressive attacks

(Winnicott, 1948). With this in mind, the child will intrinsicallybe motivated to make · reparations with the mother at times when aggressive tendencies are displayed

(Winnicott, 1948). Anotherform of anxietyexperienced by the child may result from being insecurely held, or a failure on the part of the mother to give continuous· life support through adequate mothering (Winnicott, 1952). Should these failuresoc cur, the infant's wishes and fearsin relation to the parent may be disturbed, while the infant can · also experiencepro longed excitation of the physical body, resulting in the·manifestation of defensiveprocesses (Winni cott, 1931).

17 Winnicottcl earlyoutlin ed adequate .mate��t] ��re :versus inadeq�ate matern3:l .. .

care. He first coined the term,"pri maryma ternal preo�cupation," which describes a

psychological state of heightened sensitivity of the mother to the needs of the child

e m e e a a e e e (Winnicott, 1960). Th oth r will b psychologic lly motiv t . d to.t nd to h r child's . , . .. � ' '

needs and provide "good enough care" of the infant throughprot ecting th� infant,

attending to the infant's physiological needs, providing routine care_ forthe inf�t, a�d

closely followingchang es in the infant's development (Win�icott, 19-60)._ Through t is. � :··. relationship, the infantwill learnto depend on help fro� the externalW':)rld; develop a . - . _ . .

sense of external reality; develop an integrated and stable personality based _on

interactions with the mother; and begin to have a sens� of his/he� own bo_dy (Winnic,ott,_

1935). As the mother genuinely and consistently mirrors or reflectsb ack th�_infant's

internal states to the infant, he/she will then develop an intern�lized c��cept of s�lf and

other (Wmnicott, 1971 ). The interactions between mother and child _then lead to

integration of the self, personalization, and object relating (t�e at?ilityto relate to pthers : .·. . .

both internallyand externally) (Winnicott, 197J ). Not only does the child develop a ··. .- ·. · · sense of htm/herself as an individual, the sense of mother a;s- w1ique &nd �eparate also.

becomes partof the �hild's internalund erstandings. This inte�al unders��ding, or

internalized good object then influences the child's perception pf s_elf, othe!, and wor,d.

"Inadequate mothering," or maternal care which is not attuned to the infant's ..

needs,will prohibit the internalization of self and other (1952). The jn_fant will . _ · -,.' . .. � . . . . - . . . · .

experience prolonged excitation due to not h�ving his/her _ne�ds met, a�d will_n�t r�ceiv� .

adequate mirroring from the mother, resulting in feelin_g_s o.f the self as unii:ite�ated .

18 (fragmentedsense of self),disintegrated (self as fa1ling apart), depersona1iz�� (self is ·. ' .. develop aninadequate relationship betwee1:1min� apd body, or experien�e not real), ' . . . . a .... , . .. · . .

shiftin consciousness away from the disturbing ev(?nt(1 952)'. Such d_istor,ti_o� prevent_ . ·. ' . . . . ' . . :. ·� - . the adequatedevelopment of a mature ego and force th_e chilp t� face_unthinkable ar,xiety

(1962). Thesedisturbing sensations-of intolerable anxiety._or depression in the child, . create the development of the "manic defense"as a means�o �3:rd offthe negative .

. emotions (1935). The manic defensecan include denial of in.!C?1�J reality (a lack of_.

inhabiting oneself),flight from e?{te�al reality ( daydream�ng, m��pul_atio�� of sexuality

or sensuality), suspended animation (attempts to control.. ' the'. "bad" internalparental . � ...... � ...... , objects, which will translate into disruptedrea] world relationships) (1935). Winnicott ...... ·. argued that trauma would involve deprivation or loss ofthe "g� enoughmother'� �ith ' • 1 • • . •. . • . • • theconseq uences being typically unconscious and carrying_im lications for later . " J? ...... :- ' ...... ,. psychopathology such as schizophrenia, falseself , �n� sc�izoid �ctio�ing ( 1962.)�·rThe

infant's ."true self''or sense_ of personality may be buriedbene�th a "false,se lf,')otherwise

known as the self that develops to compliantlyreact to.an environment not geared.. to meet . , ... •,' ,• .. ' ... . '. . .. the infant's needs. Thoseinfants who must adaptively_develop_a false self may fee1

inadequate or empty until the trueself may later be revealed in _!he context of a nurturing . . .· ,: . . ·,. and supportive environment ( 1960).

A Comparison of Attachment Theoryand Winnicott's Object Relation_s Theory

Fonagy states that Bowlby and Winnicottcoi ncide i� �ml?ha5izjngJh� impqrtance , . . •. • , ,'I • . . · .· of adequate caregiving to the infantas fundamentalto later development of internal

19 structures (1999). According to Fonagy (2001), both Winnicott and Bowlby stress the importance of maternalsensitivity to theinfant 's behavioral cues in the earliest years, and the degreeof sensitivity and responsiveness that will later impact psychological growthin the child. These processes occur according to both theories in the formof: a) the caregiver not impinging upon the child throughcurtailing !he child's natural expressions; b) the caregiver acting as a tension regulator forthe child; and c) the child developing an internalrepresentation of the relationship with the mother, which is the prototype for future relationships and consequent relationship behaviors (Fonagy, 2001). Cassidy

( 1998) noted commonalities between attachment and object relations theories in they both state that a) parents are powerful influenceson children, b).children develop mental models which are basedon experiences with parentsand those models will later influence children's expectations, feelings, and behavior with others, and c) experiences related to st--,,aration fromand loss of an attachment figurepJ� y a key role in the activation of these models. Both theories emphasize the importance of mother's availability to the child as encouraging the child's exploration of the world in th¥ formof secure base for attachment theory and internalized object forobject relational theory (Cassidy, 1998). Both perspectives describe the role of mother as a source of comfort in times of trouble and describe the anxiety caused by the absence of mother while asserting the development of pathology as reiated to disturbances in early experiences with the caregiver (Cassidy,

1998).

As Fonagy and Cassidy have discussed, there are striking similarities between attachment theory and object relations theory� as outlined by Winnicott. Both theorists

20 emphasize the importanceof adequate maternalcare to the infant. Adequate care of the -

infant in the earlyyears will then result in internalized psychological concepts such as

internal working modelsand an internalizedrepresentation of mother. Both theories

argue that there will be inherent fear in losing the attachment or love object� while

prolonged excitation of the infantmay carry deleterious consequences. Prolonged excitation of the infant may-result fromdisruption in care giving throughinadequate maternalcare, or loss of/separation fromthe caregiver. These disruptionsin the relationship with the primarycaregiver will cause the infant to-engage in adaptive defensiveprocesses to manage the intolerable level of anxiety that results from such disruptions. -For Bowlby, the infantwill use processes of cognitivedistortion - such as repression, cognitivedisconnection, redirection, or preoccupation whereas Winnicott emphasizes more affective disruptionssuch as feelingsof unintegration, disintegration, depersonalization, as·well as more manic defensesagainst �iety or depressioq such as denial of internalreality, flightfrom external reality, or suspended animation.

Meanwhile, Wim1icott speaks specificallyto the development of a compliant falseself in place of the authentic true self. While both Bowlby and Winnicott speak to these adaptations as carrying implications forthe development of later pathology, Winnicott appears to have been more specificin definingthe compromises made to the affective experience of self. Regardless, there has been little empirical support forhis theoretical framework.

A parsimonious integration of these two theories would look very similar to the . , comparison made between classical psychoanalytic and attachment theory. Again, the

21 emphasis is placed on p�rent child interactions, leading to th� internalizationof thec�re giving relationship. This internalizationcan be hi!ldered_thro1:1gh-,inadequate ,parepting or , loss of the caregiver. In order forthe infantto manage theresulting physiological state of fear or anxiety, defensive processesare employed leading �0-.:3 ,disto11ed exp�rience of self and other with complications for later pathology.

. :.. ;

Additional Object Relations Theories .,.· . ..

While Donald Winnicott may be considered a key <;ontributorto the lnd�p�ndent .

British Object Relations theorists, Melanie Klein is oftenassociated with the Classic

British Object Relations theories due to her expansion on the original w,orks of Sigmund

Freud. While Bowlby's training andexperience in the British-Psychoanalytic Society­ was predominantly Kleinian, he continued to emphasize th� .n�1/d. for scientific ._ .. . - · , approaches to understanding the theoretical underpinnings o{ human development

(Fonagy, 1999). MelanieKlein developed her theoriesout of her extensive work _with children and adults in therapeutic settings. She postulated the infantis subject to internal persecutoryanxieties as a result of guilt and fearassociated ,with the infant's ,inherent greedand aggression that is directed towards the mother's br�ast in the first three :to four months of life (Klein,1952). 'Ibe infant begins to develop _an internalfantasy world of .

. "good" and "bad" partobjects based upon his or her projections _of gratifyingand ....·, frustratingexperiences onto the breast and consequent introjections or taking in of those experiences into the internalworld (Klein, 1_ 952). ln this sense, the infant's _developing world of actual and whole objects in the environme�t (i.e .. ,.moth�r) .is bound _up withhis

22 or her fantasies of the experiences with her breast (Klein, 1952). The process of projection then carries the perpetuationof persecutory anxiety (Klein, 1952). As the infant desires to. greedilydevour the mother's breast, the infant fears that he or she will be greedily devoured, hence the development of the paranoid-schizoid position (Klein,

1952). Relation to partobject (breast) and splitting of idealized (good breast) and persecuted (bad breast) experiences and/or fantasiesare hallmarksof thisposition (Klein,

1952). Psychological defensessuch as splitting, omnipotence, and denial wilJ be employed by the infant to manage anxietyuntil more sophisticated processes develop

(Klein, 1952). Integration, rather thanvacillation, of destructive and loving impulses towards the breast give rise to the "depressive position," where the infantmay experience guilt, sadness, and .the desire to make reparations to the injuredloved object or good breast (Klein, 1952). The abilities to experience ambiva�ence ap.d makerepar-at ions a�e more highly developed proc�sses, which also gives the infant a more realistic understanding of the externalworld and others, and modulates persecutory anxiety

(Klein, 1952). An inability to overcome primitive defensesagainst persecutory anxiety may result in rigid barriers between the conscious and unconscious, excessive repression, and disturbed psychological development (Klein, 1952).

Fonagy (1999) has noted points of convergence between Bowlby's attachment theory andKlein 's Object relations theory in terms of stressing the importance of the first year oflife andalso noting Klein's theoretical constructs and processes appearto be operationalized by the scoring system of the Adult Attachment Interview (AAI). For · example, splitting takes the formof memory discrepancies, }ability of mental

23 · representations are. captur�d by narrative in�onsistencies., t4e· "depressive position" appears to refl�c.t secure attachmentand the ability to recognize·and accept imperfections of the caregiver, and the interview is structured to capture defensivep rocesses such as denial or idealization. Fonagy {1999) linked Klein's concept of projectiveidentifi cation to research on children withdisorganized attachment styles, notably children with histories of parents who have unresolved trauma, maltreatment, maternal depression, and prenatal parental substance abuse. He theorized that children �e forcedto_ internalize confusingand hostile aspects of their c�egivers, which .they must then externalizein order to maintain some degree of coherent self (F onagy, 1999). Meanwhile, Muir ( 1995) · has explored the possibility that Klein's theories have si�ificant overlap·with· child · attachment research and has put forththe idea thatthe Kleinian conception of infant . · · development preparesthe way foroptim�1 development of the attachmentsystem. · ·.

Attachmentmechanisms then secure the development of �arlyobject relations (Muir,.- . · 1995)

Fonagy stated that Bowlby's main objection to Kleinian theory was the neglect of actual experience, the assumption that anxiety arises fromconstitutional tendencies,. and attributing intentionalio/to the infant (1999). Attachment theory does not incorporate a death instinct and is more categorical in separating attachment styles, rather than . djmensional in Klein's universal positions (Fonagy, 1999). Jude Cassidy (1988) has consistent difficultywith Klein's theories. In particularKlein 's coreemphasis on the child's unconscious fantasiesand intrapsychic conflictto the exclusion of actual · .-· experiences, andher emphasis on early experiences withthe mother -being colored by

24 frustration, anger, and suspicion. Cassidy (1998) also argued that fromBowlby 's attachment theory, infantsare biologically predisposed to adapt to the environment, rather than being drivenby a death instinct, sex or aggression.

Heinz Kohut offered an American version of psychodynamictheory through his development of self-psychology. Kohut theori?:edthat the infantis motivated by a need to maintain a cohesive self, fueledby healthy narcissism or self-love (Kohut, 1977).

According to Kohut, a child's self begins to develop at around _18 months and at that time· significantothers, called "selfobjects,"must respond to the child's narcissistic needs in order to facilitatehealthy development (Kohut, 1977). During growth, the child's grandiose self requires confinnatory and validating experiences fromthe selfobje:ct� as well as balanced boundary and limit setting which will frustratethe child (Kohut, 1977).

As· the child continues to develop, he/she will require the opportunityto idealize the selfobjectin order to experience and participate in perceived strength, stability, calmness,-· and reintegrationof the self in times of fragmentation(Kohut, 1977). Lastly, the child will engage in twinship relations with others and focuson identificationand similarities which will later lead to differentiation (Kohut, 1977). As the child deals with normative frustrationsin theenvironment, transmuting internalizations occur where gratifying responses of the mother are internalizedinto the child's internalpsychological world

(Kohut, 1977). These internalizations allow the child to internalize productive functions that the parent originallyperf ormed(Kohut, 1977). Pervasive empathic failureson the part of the selfobjectwill hinder fulldevelopment of a cohesive self and facilitate disturbances in self-esteem regulation, hypersensitivity to failure, criticism� rebuffs, and

25 disappointments, and result 'in selves that are. fragmented;ove rstimulated,' understimulated, or overburdened (Kohut, 1977)� Such impairments can significantly impact an individual's character style, and shape personalities that are constantlyJoo kii1g· · · forselfobjects to be mirroror be adiniring, seeking selfobjects·that can· merge· to provide calming functions, or personalities that shun soci�l contadto'nairie a few (Kohut,' 1977).

F on�gy illustrated thatsimilar to Bowlby, Kohut replaced the drives of classical analysis with relational entities, requiring the attuned responsiveness· from a significant ··. other in order to develop a cohesive self (1999). According to-Fonagy, Kohut's theory· of ·· internalizationsis similar to Bowlby's concept of internalworking models, where for·

Kohut, representation of the child as omnipotent andthe parent as perfect can ·become fixed, detrimental psychological structures (1999). Me�nwhile differences take the form of Kohut empha·sizing self cohesion rather than a biologically determined relatioriship. .. pattern,the factthat Kohut places self-disintegrationanxiety above anx:iety.'about- object· · loss, and Kohut's focus on concepts such as self and narcissism which are not addressed by attachment theory (Fonagy, 1999).

Lastly, Harry Stack Sullivan was a major contributor ·to the interpersonal approach, which then laid the foundation forrelati onal/inter�ubjective theories.

According to Sullivan, an infantis bornwith expansive huma1i'·potentialities that later . . . ..� develop into a self-system, the part of the personality arising from significantinter actions with others (1954). Sullivan saw the infant as strivingto use the system to defendagainst anxiety and stressed the importance of actual two-person-exchanges· in this process (1954). He argued that earlyproblematic experiences with the.parents� such asstrict ·

26 toilet training, can significantlyimpact average human development ( 1954 ). He also . . ' . stressed the importanceof imaginary others and fantasized characteristics of the other . . . . that significantly influencethe individual andcalled thisphenomena "parataxic

distortions"(1954). Sullivan did not emphasize the importanceof intrapsychic conflict,

unconscious fantasyworld, or even motivational drives so much as he stressed the importance ofactual interactions between people and the abi1ity to observe,even I . . - •• . .- , . participatein those interactions.

Fonagy discussed overlap betwe�n the two theories in that a) both Bo:wlby and Sullivanemphasized the importanceof early parent-child interactions and both take a . : :.•. . ,, . relationalapp roach underlying human development and b) both argue fora representational system which ...is created out of the relationshipwith the primary caregivet, ,with-parataxic distortionslooking very similar to �o�l�y's_ concept? f internal working model (1999). Mauricio Cortina(200 1) concurredthat both Bowlby and

Sullivan conceptualized earlyinterper sonal experiences with the caregiver as formative . . � .. and focused on Sullivan's use of personifications,tolerably acc�ate portrayalsof interactions with primary caregivers. She also stated both tht?oristsdevote � similar importanceto defensiveoperations, such as the role of selective exclusion according to_ . . Sullivan andsegregated systems in Bowlby's work (200 I). The primary distinction between attachment theory and interpersonal theory appears to be the lack of interpersonalemphasis' on biQlogical drives and need to physical contact> proximity, and ' ... ,, . .. �> ,, , . . comfortas motivating behavior (Cortina, 200 l; Fonagy, 1999).

27 In conclusion of this theoretical di�cussion, Fonagy has clearly outlined how psychoanalysis can continue to clearly contribute to attachment theory. He argued that attachment theory could expand upon ta.icinginto consideration systematic distortions of the child's perception of the externalworld to include the fact that different infantsmay encode actual experiences in unique way� due to fantasies,internal affects, and conflicts

(1999). This line of reasoning also echoes the positions of most.object relations theorists who focusmore on adults' mental representations of se�f and others in close relationships, which are believed to be an extension of the early relationships with parents

(Levy, Blatt & Shaver, 1988). Also from a developmentfilline of reasoning, internal working models may be in conflictand/or exist in some type of hierarchy with varying .

degreesof consciousness (Fonagy, 1999).

The internalworking mode]s may be more mapeable than -origina11y theorized in

re]ation to the child's maturingrepresentational system (Fonagy, 1999). Yet another

developmental issue includes what appears to be the categorical and fixed nature of

attachment classification, without considering how the environment may impact

attachment status, particularly at certain ages (Fonagy, 1999). For example, Levy, Blatt,

and Shaver have recognizedthe typicalAAI study fin�sproportions of secure adults to be

around 65% of a sample (1988). This maybe due to the factthat theAAI was designedto

make predictions based upon the Ainsworth StrangeSit uation classification system,

whk•.h typically found 65% of American middle-class children to be secure in study

populations (Levy, et al. 1988). Meanwhile, in a twenty year longitudinal study

completed by Waters, Merrick, Treboux, Crowell, & Albersheim (2000), the authors

28 found 72% of the infantsreceived the same attachment classificationin earlyadulthood�

Negative life events or environmental factors, such asloss of parent, parental·divorce�

life-threatening illness ofparent or child, parentalpsych iatric disorder, or physical or sexual abuse by a familymember were important factorsin 44% of the cases that changed classificationcategories (Waters, et al., 2000). While psychoanalytictheorists may be very attuned to environmental influences or factorsthat can impact anindividual negatively, they also appear more comfortablein recognizing behavior patterns,both normative and maladaptiye, whi�e attachment theoristsstill conceptualize attachment status more as theoretical constructs(Fonagy,' 1999). This translates· into the rich psychoanalyticunderstanding of psychopathplogy, which remains to be explored by attachment theory (Fonagy, 1999)

Currently, researchin the arenas of psychoanalytictheory or object relations theory hasbeen next to impossible due to an inabilityto scientifically measure unconscious processes or internalized object relations. While rese�chers have employed the use of projective or semi projective measures, such as the R9rschach Inkblot Test or

The Thematic Apperception Test (TAT), such literature has been met with mixed criticism due to a lack of empirically validated and consensually agreedupon scoring· systems.; .Projective measures involve the presentation of &mbiguous, or semi ambiguous . stimuli (i.e.; a vague pictorial depiction of people) withthe intention of tapping an individual '-s unconscious processes through 1):lestories told in response to tl}.eimage.

AlthoughBowlby significantlycontributed to the fieldof psychQ}ogy through emphasizing attachment styles in children, little research has been completed which

29 actually· attempts to measure internalworking models. The majority of research relates a child's attachment style to various emotional and behavioral outcomes without ever

...... being able to document how the internal workingmodel serves to influence the child in reaching the outcome. To provide some examples, ·a brief review ofreseatch ori child attachment will be discussed next, followedby adultattachnient and object telaf101is research.

Research on Children's Attachment

The following articles neatly exemplifythe important role of maternal care in relation to infant security and later child emotional andbehavioral development.. In a , study of 129 Israeli infantmother dyads, mothers . maternal responsiveness wa; · significantly related to infant attachment status as indicated by the Ainsworth -s·trange

Situation (Koren, Oppenheim, Dolev, Sher, & Etzion-Carass·o; 2002).' TheAins·worth Strange Situation is a frequently used laboratory observation·pro·cedure·designed fo. place·· infantsinto either a secure or insecure attachment status based upon behavior displayed '· when separated and reunited with niother(Koren et at, 2002). · The Stra1ige:Situ�tiori has been found to be bothhi ghly reliable and valid (Koren et al., 2002). An analysis of · . variance revealed that mothers rated as more insightfulof their infants·' internal experiences were rated as more sensitive in their attunement to their infants' needs and · were more likely to have securely attached infants than were' mothers· not classified as . insightful, as measured by the Insightfulness Assessment. (Koren ·et aJ ., 2002). The

InsightfulnessAssessment is a series of interview questions 'presented to the mother .

30 asking her to make observations on her own videotaped interactions with her infant

(Koren et al., 2002). Such findings hold across cultures as demonstrated by Pasoda,

Jacobs, Richmond, Carbowell, Alzate, Bustamante, & Quicino (2002) in their study of 60

US and 61 Colombian mother-infantdyads. They used the MaternalBehav ior Q Sort, an observational method with an inter observer reliability of .83 and the Strange Situation..

Maternal sensitivitywas definedin terms of sensitivity in responding to infants'signals, ability of the mother to take the infantinto consideration, the positive emotional·tone directed toward the infant, an_d the ability to. cooperate with the infant. It was significantlycorr elated with infant security status,indicating that mothers higher in maternal sensitivity were more likely to have infants with secure attachment-styles than mothers lower in maternal sensitivity (Pasoda et al., 2002). This findinghe ld across cultures (Pasoda et al.; 2002).

Research on 94 preschool age children demonstrated through use of the

Attachment Q Set and teacher and peer ratings that boys and girls less secure in their attachments will show increased anger and aggression in a preschool setting, while boys who are securely attached are more popular in school (DeMulder, Denham, Schmidt, &

Mitchell, 2000). A longitudinal _study designed to measure emotion regulation abilities of six year old boys, by observing their behaviors on frustratingtasks, indicated that securelyattached boys were more likely to engage in emotion regulating behaviors such as self distraction, waiting quietly, and asking questions about the tasks thaninsecure ly attached boys (Gilliom, Shaw, Beck, Schonberg, & Lukon, 2002).

31 A seven-year, longitudinal stud)'of 1�� intematioi:,.�lly ac;lopte� c�il�e1,1_ demonstrated that higher quality of the child-motherrelationships in t�nn�of attachme�t . security (as measured by the Strange Situation)and maternal sensitivity(assessed through behavioral observation) predicted be�er soc:�a! qe�t:lol-?ment� �fsO co_n�ol� and : cognitivedevelopment in 7 year old chpdren; �s nw�ure� by �ta��ar4i�ed _scales �uq� as

IQ tests, the Child Behavior Checklist and the Teacher's.Report, to name a few (Starns,

Juffer, & van ljzendoom, 2002).

Further research with 41 children be�een tpe �ges ?t2-?-6 yea�s demo°:strated that secure attachmentto the mother correlafep with overaJl emotional understanding l I f 't � ' . , . • •

(Laib)� & Thompson, 1998). Meanwhile a hierarchicaJ regr�s�ion ind��a!ed that s�cure .. attachment predictedincreased understanding of negative emotions. in securely attached children vs. insecurely attached c�ildren (La�ble & Th�µt:psor.1_, 1_�98). . _This_ paiti�ular study measured' child attachmentthrough the percept�on pf th� mother a� she responded to theChild AttachmentQ-Set, 90 descriptive statements pfthe chi1d' s behavior, which , ' . . ,. . . has been demonstrated to be psychometrically sound. Emotional understanding, on the .,. �. •' ., 1_ • other hand, was assessed through perspective takingtasks such determiningthe affect acted out by puppets (Laible & Thompson, 1998).

Additionally a longitudinal study of 108 French c�il

observations, standardized tests, and self reports indicated that of behavioral . 8 year old . . , ' , . . securely attached children demonstrated better communication_, cognitiveengagement . and mastery motivation than insecurely attached children who sc_ored lowest on mastery -, ' . . . . motivation (Moss & St. Laurent, 2001). In a review, Ly�ops-Rutl). (1996) summarized ' ... ,' ... ; .

32 attachment related-studies where disorganized attachment--a formof insecure attachment where the child appears frightenedof the caregiver and displays disturbed behaviors in the presence of the caregiver--predicted aggressionin school-age children.

To exemplify,25 preschool boys referred to ·a psychiatric clinic due to disruptive behavior disorder werematched with normallyfunctioning boys and were rated on attachment status through behavioral observation (DyKlyen; 1996). Analysis·revealed thatattachment status of the boys was associated with the clinic status (DyKlyen; 1996) . .

This research reviewed the importanceof maternal responsiveness to infant security attachment, and traces both positive and negative child emotional and behavioral outcomes as related to the attachment sta�s, �easured in a varietyof ways. The research also emphasized cross..culturalunderstandings and the importance of attachment status, from infancyto school age. According to the theoriesof Bowlby, Freud, and Winnicott, one could arguethat these well a4apted children with secure attachment styles have functioning internal workingmodels, low conflictamong intrapsychic structuressuch as id, ego, and superego, and positive internalized object representationsof their mothers.

But little research to date has �en able to document such complex, and often unconscious psychological processes. Simultaneously, one could argue that children who are insecurely attached_most likely form maladaptive internalwo rking_ models, demonstrate high conflict among id, ego, and superego, andhave internalized negative object representations of their mothers. Only recently, with the development of more complex and sophisticated instruments, have researchers been able to empirically begin to tap unconscious processes in adults.

33 Researchon Adult Attachment Status and Object Relations : . The development of the Adult Attachment Inten'te�(AAI) by ··t\,fain.and . colleagues has proven to be a valid and reliable interview��thod to a:S-s�ss ·adult's attachment style as a function of their internalworking model (Fonagy et al., 1996). The AAI is now recognized as the "gold standard" formeasuring attachment\� ad�its· (Fonagy et al., 1996). In the AAI scoring system, there are four major· adult attachment . categories: 1) autonomous (valuing and coherent of attachment);'2) disniissive .

(idealizing, derogatory, and.cut offfrom attachment experiences); 3) preoccupied .

(passive, angry, and entangled by past experiences); 4) and unresolved (unresolved and/ordisorganized concerningtraumatic or loss attachment ·experience·s) (F��agy, et al., · ,

1996). The interview method encourages su�jects 'to conj�e and ·a�;s�ss -�ttachnient related autobiographical memories in order to· assess the i�dividuai's curr�nt working ° model of attachment relationships (Bakersman-Kranenburg & �an. IJ�e�doo�,' I 99 3 f Eighty-three mothers-participated in a project �esigned to study the discriminant validity and reliability of the A.AI (Bakersman-Kranenburg & van IJ�endoom, 1993). Results indicated the AAI was independent of non-attachment r�lated ·autobiographical me�d�y, verbal and performance intelligence, and ·social desirabiiity, while· AAiclassifi�ations were reliable over a two month period (Bakersman-Kranenburg& van· Uzend6�m, 1993.). .. The development of the AA_I has provided an opportu�ity tostudy aduit atta�hment �tyles, .

. ,· with a preliminary emphasis on relation to children's attachment st��' 'clinical ... diagnoses, and treatment.

34 Slade, Belsky, Aber, & Phelps (1999) explored.the relationships between

. . mothers' representations of their relationships with their children, using theParent

Development Interview, adult representation of attachment, andobserved mothering.

These researchers determined that autonomous mothers represented their relationships

with toddlers in ways that were more coherent and conveyed more joy and pleasure in the

relationship than did dismissing and preoccupied mothers, while dismissing mothers . . - expressed more'anger in the Parent Development Interview (1999). The researchers

concluded that mother's state of mind regarding attachment and her own attachment experiences related to theway she perceived her relationship with her child and could

influence actual parentingtechn iques (1999). Additional researchhas demonstrated that

mothers' attachment status and-infantsecurity were associated with emotional availability scores as measuredby behavioral observationsof the mother-infantdyad (Aviezer, Sagi,

Joels, & Ziv, 1999). Results indicated that security of.infant attachmentand mothers'

autonomous representations of attachment were associated with higher emotional availability scores, while poorer emotional availability was observedin dyads of insecurely attached infant<; and mothers classifiedas insecure 'regardingattachmen t status, thus providing an argument that adult attachment status will behaviorally influence

parenting and consequent child attachment status (Aviezer, et aJ., 1999).

George & Solomon explored experiences of caregiving, adult attachment status, and child attachment status and determined significantconcordance bet\veen ratingsof maternal representations of caregiving and child attachment, concordance between

·mother's currentrep resentation of attachment and her representation of caregiving, and

35 concordance between adult attachment classi_ficationsand child attachment classifications( 1996). Autonomous mothers who had a maternal representation fora secure base had children who were securely attached, d�1ached mothers with rejecting representations had avoidant children, while enmeshed mothers who had uncertainty in their representations·had ambivalentchildren (George & Sqlomon, 1996). Further evidence was provided by Cowan, Cohn, Cowan,& rearson (1996) who found, in a study of twenty-seven mothers and twenty-s�ven fathers,that aspects of the adult attachment status of the parents acted as s!rong moderators in predicting teachers' descriptions of children's internalizingand externalizingbehaviors. More specifically, men's lovingand anger ratings on the AAI were associated with their parentingstyle and child's consequent level of externalizingbehavior, w�ile women's loving and anger scores on the AAI were associated with parenting style and child's consequent level of internalizingbehavior (Cowan, et al., 1996).

Adult attachment status has ·not only been linked to parenting styles and child behavior outcomes, but also to adult psychopatµology. Pianta, Egeland, & Adam studied adult attachment classifications in relation to symptoms reportedon the Minnesota

Muliphasic Personality Inventory-2 and found that dismissing mothers reported little psychiatric distress and had the lowest scores on anxiety, while preoccupied mothers scored higheston indices of self-perceived distress and relationship problems (1996).

Fonagy et al. (1996) compared the patternsof attachment and psychiatricstatus in 82 nonpsychotic inclients and 85 case-matched controls using the AAI and foundthat psychiatric clients were more likely to be classified as preoccupied and unresolved with

36 respect to loss or abuse thannonpsychotic clients. More specifically on Axis I, anxiety was associated with an unresolved status, and AA.Iscales were able to distinguish depression and eating disorders (F onagy et al., 1996). Meanwhileon AxisII; borderline · personality disorder was related to a disorganized attachmentstyle (Fonagy et al., 1996).

Individuals rated as dismissing of attachmentwere more likely to show improvementsin · · psychotherapythan those rated as preoccupied (Fonagy et al., 1996). Such.research· implicitly sets the stage forfuture research in thearea of object relations, since object relations theorists argue that psychopathology js intrinsically indicative of disturbed or distorted object relations.

Diamond, Clarkin, Levine, Levy, Foelsch, & Yeomans (1999) were also able to document in their preliminarywork with two subjects, that borderline personality disorder tends to carryfeatures of an insecureand disorganized attachment style, with poorer object relations being intimately tied to symptomololgy of the insecure attachment. With the aid of Transference-Focused Psychotherapy (TFP), an individual's insecure status can shift to autonomous/secure, as measured by the AAI(Diamond et al.,

1999). These two preliminary studies are the only work to date incorporatingattachment theory and research methods to furthera developmental understandingof ·· psychopathology and clinical implications. Research addressing .clinical implications inherently tap object relational processes, but empirical evidence for the role of object relations is still significantlylacking. Diamond ·et al. ( 1999) conjecturedthat an individual's images of self and others, as measured through object relations inventories, are derived fromearly attachment history andplay a role in the internal, working models

37 as manifested through significant interactions with others. Although the research could . not draw conclusive findingsdue to the small sample size, the researchers con.eludedthat · - insecure attachment categories may encompass individuals with different levels of object relatio�s development, and that additional measures of the level and quality of self.;.- and · . object representations may provide relevant informationregarding therapeutic. .· interventions (Diamond, et al., 1999).

Levy, Blatt, and Shaver ( 1988) stated that overall, ·adult attachmentresearchers · · ·· :,·,, have not investigated or incorporatedthe ideas and measures developed by object relations theorists. Theseresearchers attempted to document a.clear relationship between attachmenttheory and object relations through exploring undergraduates'- mental · - representations of their parents throughthe use of attachment self-report.inventories and thematic content and structure of parental description narratives. They foundthat securely attached participantshad representations of parentsthat .were differentiated, elaborated, benevolent, and less punitive. Anxious-ambivfllent participantshad·· representations that were benevolent, but also punitive. Dismissing participants described their parents with less differentiation, more punitiveness andmalevole nce, whereas fearfulparticipants also characterized representations as pmiitive and .· . � � . malevolent, but with clearer differentiation and complexity. Levy, B1att, and Shaver concluded that understanding the internalized parental representatioriSassociated with each attachment style contribute to increased knowledge ofsocial behavior (1988). ·: ·

Previous research has demonstratedthat objec:t relations deficitscan -distinguish individuals with borderline personality disorder froinschizoaff ective ·and schizophrenic ·

38 groups (Bell, Billington, Cicchetti, and Gibbons, 1988). It has also been foundthat

individuals with borderline personality disorder experiencethe object relational world in

a way that is different fromthe perception of normals and depressives· (Stuart, Westen,

Lohr, Benjamin, Becker, Voros, and Silk, 1'990). Tramanto, Javier, & Colon (2003) also

foundthat object relations can ·discernindividuals diagnosedwit h borderline personality

disorder fromcontrols, as well as delineate various subgroups of borderline personality

disorder. Through use of various-object relations measures these researcherfound evidence that the internal_world of an individual ·with borderline personality disorder is highly malevolent, while providing documentation forthree subgroups moving away

(those who may have more engulfmentfears), moving toward (those who seek a fantasy

external object to contain them, and moving against (those who may experience more anger and rage) (Tramantoet al., 2003 ).

Further, the newer empirically validated Social Cognitionand Object Relations

Scale (SCORS), to be used when rating an individual's Thematic Apperception Test

(TAT) stories, can successfullydistinguish more pathological responses in the TAT stories, differentiatingborderlines, major depressives, and normals by illustratingtheir object relations (Westen, Lohr, Silk, Gold, and Kerber, 1990). 'The SCORS has been· determined to have adequate inter-rater reliability and construct validity, with The

Construct of the Object measure typically used with the Rorschach inkblot test (Hibbard,

Hilsenroth,Hibbard, & Nash, 1995) The SCORS was originally designed to tap object relations through measuring unconscious representations of people, emotional tones of relationships with people, and self-concepts (Westen et al., 1990). Westen initially

39 argued that the TAT was a good instrument formeasur�ng object relations due to social aspects of the stimuli which allows access to cognitive and affective motivational patterns. to emerge in narrative response to the cards (Westen, 1991). ·Westen designed his scale to reflectobject relational theories of representation of people in relationships, affects, conflicts, attributionsof caus�s of behaviors, and. capacity forinvesting in relationshipsas implied by Melanie Klein, Otto. Kemberg, and W.R .D Fairbairn (Westen, 1991). Currentlythe SCORS has progressed into an eight dimension scale including:

Complexity of Representation of People, Affective.Quality of Representations,

Emotional Investment in Relationships, Emotional Investment in Values and Moral

Standards, Understanding of Social Causality, Experience and Management of

Aggressive Impulses, Self-Esteem, and Identity and Coherence of Self on a seven point likertsystem, with lower scores indicating poorer o�ject relations (Westen, 1995).

Ackerman, Clemence, Weatherill, & Hil�enroth ( 1999) used the updated version of the

SCORS to successfullydi fferentiateamong individuals diagnosedwith antisocial personality disorder, narcissistic personality disorder, borderline personality disorder and

Cluster C personality disorder. They were able to achieve Spearman�Browninterrater reliability rates of .74 or above (with the exception of I�entity and-Coherence of Self) ·

(Ackerman, et al., 1999). To date, no research has ad�ressed the relationship between adult attachment status and object relations as µieasured�y the SCORS.

This dissertation theoretically explored similarities and dissimi]aritiesamong attachment theory, classical psychoanalytic theory, and object relations. -An integration of these theorieswould emphasize earlyparent child interactions as fundamental to the

40 internalizationof cognitive and affectiverepresentations of thecaregiver and the nature of the caregiving relationship, in order to formunderstandings of self and others.

Internalizationscan be hindered, distorted, or disrupted throughinadequate parentingor traumatic experiences. In order forthe infantto manage the resulting physiological state of fearor anxiety, defensivecognitive and affoctiveprocesses are used which oftenresult in distorted experiences of self and other, creating implications forlater pathology. A weakness in attachmenttheory involves a clear definitionof internalworking models.

How do these unconscious processes and defen�ive patternsserve to shape an individual's personality and psychological functioningin daily experience and/or relationships with others? Most attachmentresearch relates the attachment status to some other aspect of psychological functioningwithout clearly exploring the processes underlying how and why individuals have different attachment statuses and consequently differentinternal workingmodels. Only recently, with the development of the AAI and

SCORS, both designedto tap unconscious processes, is psychology at a place and time where research attempting to integrate the described theories can occur.

The AAI is highly recognized as the most effectivemeasure of adult attachment status, in lieu of self-report measures, because it a) incorporatesa fourcategory system with unresolvedstatus, instead of a typical three category system, b) taps unconscious processes to determinestatus, and c) is not subject to a social desirability factorcommon to most self-report measures. Un fortunatelythe AAI is a laborious instrumentwhich can take an hour to an hour and a half to administer, fourto six hours fortranscription, involves a complex scoring system, can be difficult to achieve interraterreliability, and

41 provides a limited understanding of psychological processes through only providing an attachment classification.

Meanwhile George, West, and Pettem ( 1999) have designeda new methodology using narratives given in response to semi projective stimuli that are designedto elicit attachment representations, known as the Adult Attachment Projective (AAP). This is the firstpro jective measure of adult attachment status and uses variables such as coherency of narrative, relatedness, agency of self, and defensiveprocessing to designate fourma jor adult classification groups: secure, dismissing, preoccupied, and unresolved

( George & West, 2001 ). The eight cards of theAAP are administered_ in about half an hour, involve a more accessible scoring system than the AAI, uses less time in transcription,and provides more informationregarding internalprocesses, such as agency of self and defensivepr ocessing. Preliminaryvalidation research has shown AAP interjudge reliability for secure vs. insecure classificationswas .93 (kappa=.73, p<.000); interjudgerel iability forthe fourma jor attachmentgroups was .86 (kappa=.79, p<:000); convergence between AAP and AAI secure vs. insecure groups was .92 (kappa=.75, p<.000); convergence forthe fourma jorAAP groups with the AAIgroups was .85

(kappa=.84, p<.000) (George & West, 2001).

Since the AAP is a relatively new instrument., little research has been completed using this measure. Currently, West & George (2002) have founda strong association between fe male dsythmic depression and the preoccupied category as measured by the

AAP.

42 This dissertation attempted to provide a clearer understandingand definition of internalworking models by relating adult attachment classificationas measuredby the

AAP with object relations representations as measured by the TAT and SCORS. Further understandingof unconscious processes which underlie adult attachment classification can be illustrated by empirically investigating how attachment statiJ.srelates to an individual's understanding of self and others' internal states,the ·affective nature of relationships, the level of emotional investment in relationships, understanding of social causality in relation to interpersonal events, and the experience and management of aggressive impulses, as measured by the SCORS. This research carries inherent impo1tance in farther illustratingconcepts of attachment theory, providing additional empirical evidence forebject relations theory, and providing a more comprehensive picture of human development with implications for understanding and treating psychopathology. This study specifically hypothesized that individuals clac;sifiedas securely attached by the AAP coding system would have significantly higher ratings on the eight SCORS dimensions of the TAT than individuals with insecure attachment . styles.

43 II. Method

Participants

100 undergraduate students were recruited fromintroductory and upper level psychology courses at the University of Tennessee, Knoxvil1e to ensure a distribution of ages. Three subjects were not used in the statistical analyses due to missing information.

Data from97 students (28 men and 69 women, mean age :;::21 years, min. = 18 years, max. = 35 years, and STD = 3.86 years) was used in the study. Ethnic backgrounds included 87 Caucasians, 7 AfricanAmericans, and 3 As�ans. Students received extra credit in their psychology courses in return fortheir involv�ment. Subjects were treated in accordance withthe "Ethical Principles of Psychologists and Code of Conduct"

(American Psychological Association, 2002). The Institutional Review Board of the university waived written infonnedconsent because there was no identifying information on the data collected and no foreseeablerisks to the subjects.

M_? terials · TheAdult Attachment Projective (AAP) is comprised of eight drawings, a neutral scene to aid the individual in becoming accustomed to the test, and seven scenes specificallydesigned to elicit attachment related responses (i.e., illness, solitude, separation, and abuse) (George & West, 2001). The primary researcher in this study attended an extensive, formal, two-week training seminar conducted by the test developer of the Adult Attachment Projective (AAP), w�ich add�essed administration, scoring, and interpretationof the AAP. The primary researcher then trained four other graduate students in the clinical psychology doctoral program over the span of a semester, meeting

44 for two hours on a weekly basis to review practice scoring. Analyzing the AAP emphasized three aspects of the story narrative:..discourse(whether or not the individual may be overwhelmedby attachment related distress to the degree that they share Personal

Experiences, and level of Coherency in the narrative, mainly measured by Quality of the plot, Quantity of the story length);story content and action, as illustrated by Agency of

Self ( degreethat self is moving psychologically or behaviorally toward integrationor understanding), Connectedness (character's desire to be with others), and Synchrony

(interactive behavior between characters); and defensiveprocessing, primarily through .

Deactivation (defensive exclusion that enables the individual to diminish, dismiss, devalue, or minimize the importanceor influenceof attachment stimuli), Cognitive

Disconnection ( defensive exclusion that splits attachment information into opposing images or story lines, thus creating confusingand distance fromattachment related stimuli), and Segregated Systems ( extreme formof defensive exclusion that cognitively segregates extremely painfulemotions associated with loss or trauma from consciousness, oftenresulting in mental dysregulation and disorganization) (George &

West, 2001).

·Prototypical patterns of the adult attachment classification system are as follows (George & West, 2001):

1. - Secure individuals demonstrate high Agency of Self in depicting characters

which have the ability to draw upon internal resources and attachm�nt figures

to address attachment relatedstress while showing the importanceof

relationships with highConnectedness, as well as providing descriptions of

45 reciprocal interactions with high Synchrpny. These individuals do not need

to rely heavily on defensiveprocesses to cope with attachment distress or

anxiety and they demonstrate high to moderate Coherence while maintaining

clear self-other boundaries as evidenced by Jow Personal Experience.

2. Dismissing individuals noticea�ly avoid or ignore direction expressions of

attachment in their narratives and will take action themselves rather than

relying upon attachment figurest� alleviate anxiety, resulting in moderate

levels of Agency of Self. These individuals tend to portraycharacters which

have low Synchrony and moderate Connectedness and will typicallyutilize

defensive Deactivation, facilitatinglower Coherency in their stories.

3. Preoccupied individuals also" do not use attachment to resolve distress, and

frequently describe characters incapableof any action at a11, resulting in low

Agency of Selt� The characters qepicted are often basically alone, low

Connectedness with poor Synchrony, and typically fall upon Cognitive

Disconnection defensiveprocesses to address anxiety. This form of

defensivenessoften results in stories which are low in Coherence and suffer

from lengthydescriptions of Personal Experience.

4. Unresolved individuals portray stories which are marked by the individual's

failureto reorganize or integrate a Segregated System in stories which are

typically low in Coherence.

The primary researcher also trained three other clinical psychology graduatestud ents in administration, scoring, and interpretation of the Thematic Apperception Test (TAT)

46 as measured by Drew Westen's Social Cognition and Object Relations Scale (SCORS)

(1995). Training lasted an entire semester, with one-hour weekly meetings to review practice scoring. The SCORS is typically based on fivecards featuring individuals alone and individuals together. Cards 1, 2, 3BM, 4, and 13MF were used, as the standard card selection fromother SCORS studies. The TAT was scored by the latest version of

SCORS (Westen, 1995), which taps eight dimensions of unconscious object relations.

Each variable is scored on a 7-point anchored rating scale ranging from 1 (indicating more pathological responses) to 7 (indicating healthier responses) (Ackem1an, et al.,

1999). The eight dimensions are as follows(Ackerman, et al., 1999; Westen, 1995):

1. Complexity of Representation of People-assessessubject 's relational boundaries

and ability to integrate both positive and negative attributes of the self and others,

including richness of the representation (Ackerman� et ·aL� 1999).

2. Affective Quality of Representations-assessessubject 's expectations fromothers

in relationships and how the client describes significant relationships(Ack erman,

et al., 1999).

3. Emotional Investment in Relationships-assesses subject's level of commitment

and emotional sharing in relationships (Ackerman, et al., 1999).

4. Emotional Investment in Values and Moral Standards-assesses subject's ability to

behave in a situation without a sense of guilt or remorse versus a willingnessto

challenge moral questions in an thoughtfulfashion (Ackerman, et al., 1999).

5. Understanding of Social Causality-assesses subject's understanding of why

people behave as they do (Ackerman, et al., 1999).

47 6. Experienceand Management of Aggressive Impulses-assess a subject's ability to

controland appropriately expression anger (Ackerman, et al., 1999).

7 � Self-Esteem-assesses affective quality of self-representation (Ackerman, et aJ.,

1999).

8. Identity and Coherencieof Self-assesses a su�jQct' s level of fragmentationand

integration (Ackerman, et al., 1999).

Subjects were administered the AAP, followed by the TAT, and lastly a

demographicquestionnaire, which included questions about loss due to death or

separation. The AAP was always administered first in accordancewith directions

outlined by the test constructor in the training seminar, since the AAP is specifically

designed to tap an individual's attachment system (Carol Qeorge, Personal

Communication, 2003). Data were collecteclin designated assessment rooms· at the

University of Tennessee Psychology Building, where tqe examiner hand documented

subjects' responses verbatim. Testing typically lasteq an hourto two hours and was

completed in one session. At the end of the data co1lection session with the subject,'the

. purposeof the study was explained and any questions were answered.

The protocols were randomly distributed to the judges fors �oring. TAT protocols were scored separately fromthe AAP protocols. When scoringth e TAT protocols,

judges were instructed to score across similar cards rather than across �ntire protocols to

prevent the halo effect. The scores foreach dimension were then averaged across the five

cards to create a single average score on each dimension foreach subject Scoring an

AAP protocol in its entirety rather than across cards is the accepted methud of scoring

48 (Carol George, Personal Communication, August 2003). To calculate reliability rates,

the five AAP raters all scored the same subsample of twenty protocols selected fromthe

data. Twenty TAT protocols were also randomly selected andscored by all fourof the

TAT raters. Inter-rater reliability among the judges was determined using multi-rater

kappa forthe AAP due to the categorical natureof the variable, while two-way mixed

intra-class correlation coefficientswere calculated forthe continuous variables on the

SCORS. Both analyses show excellent reliability rates, with most ranging from.8 1 thru

1 .96 (See Table 1, Appendix A ). After reliability analyses were completed fourprotocols

from each AAPjudge and fiveprotocols from each TAT judge were randomly selected to

· be used in the statistical analyses. The remaining protocols were randomly distributed to

the raters for completion of scoring. Raters scored an average of twenty A.AP protocols

and twenty-fiveTAT protocolsof the entire data set.

1 All tables are located in Appendix A. 49 III. Results

Final classificationof adult attachment styles revealed a disproportionate number of individuals fallinginto the four categories (secure = 44, dismissive = 11, preoccupied

= 9, and unresolved = 33, Total N = 97). Dismissive, preoccupied, and um-esolvedall reflectinsecure states of mind regarding attachment and were therefore collapsed into one group forthe purpose of more equal comparisonto the secure group (secure = 44 and insecure = 53, Total = 97). Box's Test of Equality of Covariance Matrices shows homogeneity of covariance between the independent groups (Box's M = 49 .41, F (36, 61)

= 1.25, p = .15). Levene's Test of Equality ofE�or Variances shows the dependent variables have equal group variances (See Table 2, Appendix A).

Prior to testing the hypothesis that secure individuals would score higher on the

SCORS measure of object relations than insecure individuals, demographic data were taken into consideration to eliminate the possibility of covariance. Demographic factors ,. that may influencethe dependent variables were gender, age, whether parent(s) raised the subject, loss of parent due to death or separation, parental divorce, personal inj ury or illness, and recent loss of someone close. Age was divided into three categories, students in their late teens, twenties, and thirties. Many subjects indicated loss of someone close to them, other than a parent, in the past year while other subjects reported losses that occurred up to eighteen years ago. Descriptive statistics of how many years had passed since the loss indicated the mean number of years to be 2. 74. With this information subjects were divided into twogroups, those who had a meaningfulloss in the past three years (N = 46) and those who did not experience a loss or the loss occurred over three

50 years ago (N = 51). Demographicinformation is shownin Table 3 (See Table 3,

Appendix A). Pearson's r correlations reveal that demographic factors do not significantlyrelate to the dependent dimensions of object relations (See Table 4,

Appendix A). In particular, whether a su�iect experienced a meaningful loss of someone other than a parent in the past three years or experienced parental divorce did not significantlyrelate to any of the dependent variables.

Contrary to pqpular literature, this sample has a surprisinglylarge number of individuals classifiedas insecure. To furtherilluminate this phenomenon, chi-square

analyses were completed to assess patterns of secure versus insecure distributionsin

relation to meaningfulloss of someone other than a parent, experience of parental

divorce, and year in college. Analyses show there were no significantpatterns among these categories (See Tables 5; 6, and 7, Appendix A).

Multivariate analysis ·of variance using Hotelling's Trace indicates a significant relationship betweenadult attachment status and object relations (F (8-'-91) = 3.2, p =

.()03). Descriptive statistics including mean, standard deviation, and e&timates of95% confidenceintervals are shown in Table 8 (See Table 8, Appendix A). Pairwise comparisons show a consistent patternwhere secure individuals score significantlyhigher

on the measure of object relations than insecure individuals on seven of eight SCORS

dimensions (See Table 9, Appendix A). For example, individuals classifiedas secure in

attachment status had significantlyhigher scores on Affective Quality of Representations

than ·individuals classifiedas insecure (mean difference = .521, p = .000).

51 IV. Conclusion

The original hypothesisthat secure individuals would score significantlyhigher on SCORS dimensions of object relations was supported by the results of this study. In reviewing the early works of John Bowlby, Sigmund Freud, and Donald Winnicott, there are clearly similarities in the theoretical models of attachment theory, classical psychoanalytic theory, and object re1ations. An integration of these theories would emphasize early parent-child interactions as fundamental to the internalizationof cognitiveand affective representations of caregivers and the nature of the care giving relationships. The internalizationscan be hindered, distorted, or disrupted through inadequate parentingor the experienceof trauma. TQeinfant in attempting to manage intense fear, distress or anxiety in the face of loss, separatjQn, pathogenic parenting, or - l j - trauma will use cognitive and emotional de�ensive process�s, resulting in distorted experiences of the self and other with implications for ,at�r pathology. The findingsof this study provide an empirical link between attachment theory and object relations.

These findings alsoprovide a more comprehensive understanding of John

Bowlby's concept of the internalworking model, w�ile proving the utility of an -, ' innovative new psychological instrument, the Adult Attachment Projective. Application of the Social Cognitionand Object Relations Scale with the Thematic Apperception Test in a normative population was also dem�mstrated. And lastly, additional outcomes of this

. study enabled initial explorations of other factors, such as experience of meaningful death other than a parent, parental divorce, and year'in college, which may also impact attachmenttheory and an individual's attachment system.

52 John Bowlby originally definedthe internalworking model as a largely unconscious, internalset of cognitiveand affectiveguidelines which influencean individual's understanding of self and other in the context of interpersonal relationships

(Bowlby, 1988). According to Bowlby, the internalworking model will influencethe individual in perceiving events, anticipating futureevents, shaping emotional reactions, and constructing behavioral responses (Bowlby,19 73). The internalworking model forms the individual's "state of mind" regarding attachment experiences and is inherently classifiedby measures of attachment.

According to George and West (2001), secure individuals have a "state of mind" regardingattachment where they can call upon internalresources or actual attachment figures in times of need or stress. These people value attachment relationships· and the importance of close connectionswith others while maintaining clear and appropriate boundaries. Theydo not need to rely heavily on psychological defensiveprocesses or mental distortionsof reality to manage attachment related anxiety. Meanwhile individuals classifiedas insecure, have dismissing, -preoccupied,·or unresolved states of mind regardingattachment and will have difficulty absorbing the importanceand · beneficial impact of attachment relations for various reasons. Dismissing individuals may noticeably avoid or ignoredirect expressions of attachment and focuson independence, personal strength, self-reliance, and social scripts or rules to manage

. attachment related anxiety. Preoccupied individuals frequently experience a sense of impotence in managing attachment related distress and tend to use forms of cognitive confusion,mental distortion, and contradictions as defensivemaneuvers in the faceof

53 anxiety. Lastly, individuals classified as unresolvedtypically have some type of personal trauma or loss in their history which may be psychologically unprocessed or only partiallypr ocessed, and is reflectedin their projective narratives. Research on children indicates that secure children have improved psychological health, with better emotional and behavioral functioningthan insecure children (DyKlyen, 1996; Gi1liom et al., 2002; Labile & Thompson, 1998; Moss & St. Laurent, 2001; Starneset al., 2002).

Preliminary research on adults indicate the unresolved attachment style may be related to borderline personality disorder (Diamond et al., 1999; Fonagy et al., 1996).

Until now, there has been littleempirical research on defining and measuring the internal working model. Fonagy (1999) has stressed the importance of providing further definitionto the internalworking model. He also ·indicates that attachmenttheorists present the internalworking model and classificationsof attachment as theoretical constructs, rather than observable patterns of behavior (Fonagy, 1999). In this sense, object relations andmeasures such as the SCORS has exp&ndcd beyond attachment theory in providingpsychological differentiation to the cognitiveand emotional processes of individuals experiencing differentforms of psychopathology (Bell et al., 1988; Stuart et al., 1990; Tramanto et al., 2003; Westen et al., 1990). Bowlby's theorieswere originally informed by psychoanalytic and object relations theories, but he diverged from these influencesto place more emphasis on actual physical experiences with the primary caregiver, such as separation or loss (1988). Although object relations theorists be]ieve · internalizedmental representations stem fro� ear]y caregiving relations, they place more emphasis on adults' associations, memories, and fantasies of those experiences (Levy,

54 Blatt, & Shaver, 1988). This divergence has almost become a limitation of attachment

theory; unique fantasies, individual experience of affector conflict, personal styles of

encoding information,·and systematic distortions of the external world, all of which may

influence the development of the internalworking model, do not seem to be taken into

consideration (Fonagy, 1999). The findings of this study provide preliminary evidence

that aspects of object relations can definediff erences in the internalworking model

between secure and insecure individuals.

Although it is not possible at this point in time to ascertain levels of conscious

versus unconscious influencesof object relations, rudimentary profilesof the internal

working model can be constructed when interpretingthe differences between secure and

insecure individuals. The results show that secure individuals ha:ve significantlyhigher

• scores than insecure individuals on the SCORS dimensions of Comple�ity of Representation of People, Affective Quality of Representations, Emotional hwestment in

Relationships, Understanding of.Social Causality, Experience and Management of

Aggressive Impulses, Self-Esteem, and Identity and Coherence of Self. The factthat this

project analyzed overall averagesof SCORS dimensions across five Thematic

Apperception Test cards in a young adult college sample speaks to the robustness of these

findings.

The internal working model, as defined by object relations profilesof secure

people in comparison to insecure people, would include higher scores on Complexity of

Representation of People. The secure individuals have better ability to separate self from

other, improved capacity to integrateboth positive and negative qualities of self and

55 other, as well as more enriched description of self and others' internalfeelings and thoughtprocesses. Secure people also have higher ratings on AffectiveQuality of

Representations, meaning they tend to expect and describe more positive interactions with others in relationships than insecure individuals. Secure subjects portray higher levels of commitment and sharing in Emotional Invest�ent in Relationships, as well as a broader grasp of humanmotivation and cause of events in Understanding of Social

Causality. In Experience and Management of Aggressive Impu)ses;secure people more appropriately control and express anger, while maintaining more positive affective tones of self-representation in Self-Esteem. Lastly, secure subjects experience higher levels of integrationand less :fragmentation in Identity and Coherence of Self. Overall, these findingssuggest that secure individuals have more complex, dynamic,and healthier object relations profilesthan insecure in�ividuals, with implications indicating different styles of internalworking models.

Tn ese implications are inherently important inu�derstandinghuman developmen� as well as conceptualizing and implementing treatment in psychotherapy.

With a more comprehensive grasp of the internalworking model, psychological and behavioral interventions can be more·aptly modified at any level'of treatment. Bowlby set out fivetherapeutic tasks, strongly informed by analytically oriented principles associated with the works of Fairbairn, Winnicott, Guntrip, Sullivan, and Kohut (1988).

The firstof these tasks is to provide a safeenvironment or secure base where the client can consider and explore painfulaspects of life with a trusted companion (Bowlby,

1988). Hypotheticallyinformed by the results of this study, secure individuals with

56 higher scores on Complexity of Representation of People, Affective Quality of

Representations, and Emotional Investment in Relationships may have a dearer grasp on the intention and nature of the therapist, tend to expect positive interactions with and to be helped by the therapist, while easily engaging in an invested relationship with the therapist. An insecure individual may have difficulty relating to the therapist as an individual, perhaps believing that the therapist is there to tease, harm, or violate them or considers the therapist in a role to satisfy the client's needs and desires. The client may expect negative interactions with the therapist and have difficulty trusting and engaging with him or her. With this knowledge, a clinician may be more aware of a client's specialized needs and sensitivities and cantailor responses and interventions accordingly.

The second task is to assist the individual in investigating how he or she engages others in relatiooships,explores expectations of feelings and behaviors forself and others, discovers unconscious biases, and illuminates the ways an individual contributes to the experience of difficult situations (Bowlby, 1988). Here, a secure individual with better ratings on ·Complexity of Representation of People, Affective Quality of

Representations, Emotional Investment in Relationships, and Understandingof Social

Causality would significantly differ froman insecure individual. A secure client may more easily grasp the complexities andeven conflictingnature of thought'S, feelings, and actions in self and others. The client may better understand his or her own motivations and factors which influence others, where an insecure individual may be more limited. A secure individual anticipating positive interactions may act more positively towards others, while an insecure individual may elicit negative responses from others through

57 acting in a way where the negative response is already expected. Due to maladaptive internal working models, an insecure individual could be less invested in relationships without an understanding of why or how this phenomenon is occurring. An insecure individual would also have a poorer understanding of chains of events, and may not fully realize the ramificationsof inteiTelated events, such as the contributions of their own responses and behaviors. In this situation, a therapist might be prepared to give careful attention to adequate mirroring of the client's experience with specificre-iterat ion of events to increase the client's realization of unconscious influences and personal contributions.

According to Bowlby, the third task is aiding the c1ient in examining the relationship between self and therapist, including perceptions and expectations of the therapist which may be strongly shaped by internalmodels of parents and self ( 1988).

Differences in scores on Experience and Management of Aggressive Impulses, and Self­

Esteem are salient here. Aninsecure individual may be less aware of the expectation of negative interactions with others, even with the therapist, or unaware of potentially disconnected, dissociated, repressed, or denied experiences of anger, which could be unknowingly directed towards the therapist:. Also, an irisecure individual may have increased feelings of poor self-esteem and negative perceptions of the self, or an unconscious experience of the self in a negativeway. Thus,with an insecure individual treatment may be enhanced if the therapist has heightened awareness to transference responses which pull for negative interactions with the therapist and consciously or unconsciously direct angertowards the therapist. The therapist is encouraged to actively

58 listen forstatements which indicate the client has poor self-representations, for example the client feelsunworthy of being seen by the therapist, the client feels he or she does not deserve to be seen by the therapist, the client feels offended or harmedby the therapist in some fashion.

The therapist's fourthtask is to help the client see that many current perceptions, expectations, feelings, and behaviors are products· of event� that occurred in childhood, while aiding the client in dealing with painfulemotions that arise out of this growing awareness (Bowlby, 1988). Understanding of Social Causality and Emotional Investment_· in Relationships play a vital role during these times in therapy. A secure individual may have or come to have an insightfulawareness of the impact of early childhood experiences in shaping one's personality, understanding of the world, and currentday experiences with others, whereas an insecure-individual may struggle to grasp this concept. It is oftenpainful and difficultfor clients to revis4t the problematic times in their childhoods and to accept that aspects of those painful times continue to influence them today. Strong emotional investment in the relationship with the therapist is what enables clients to undertake this arduous task. The clinician must be prepared to illustrate meaningful linksto the past as the events unfold in therapy, but in a timely fashionif the insecure individual is not able to process the therapist comments. A therapist should also be preparedthat an insecure individual is most likely insecure because of the trauma experienced in the past and should expect intense affectto manifest itself in therapy. The therapist must continue to provide.a safe, supportive, holding environment to the client.

In many ways, the strength of the therapist in sitting with, tolerating, exploring, and

59 digesting overwhelminglystrong affect with the client is what maintains the therapeutic alliance.

And lastly the fifth task of the th�rapist is to enable the client to recognize that images or models of the self and others are derived from the past,but are no longer appropriate to the present and/or future. Theclient will have better understanding of self and others andcease being imprisoned by their past unconscious influences, enabling him or her to feel,think, and act in new ways (Bowlby, 1988). According to Bowlby, all of these tasks are interrelated and many tasks intertwine in even one session (Bowlby,

1988). The previously mentioned fouriasks and corresponding measures of object relations appear to culminate in this final task. But specificto this hoped for fruition of therapy, seems to be the role of Identity and Coherence of Self. A secure individual may

already have a strong sense of identity, with less experience• in fragmentationof the self t and have the resources to connect with the therapist, engage with and exp_lore with therapist andprogress throughthese tasks of therapy in a steady fashion. Meanwhile an insecure individual may require considerably more time in establishing an alliance with the therapist, trusting the therapist, and experiencing the therapist in a realistic and· grounded fashion. Some of the time required forthis slow and deliberate process may be due to poor sense of identity, which can i:q.hibitexploring therapeutic grounds and learning from these new experiences. Even the most positive experience can be overwhelmingand disorganizing to a severely damaged individual. The c]inician should be prepared that an insecure individual is refonning their sense of identity that is

60 independent fromthe pastexperiences he or she might have had as a child, and insightfullyrespond to regressions with patience and understanding.

Theconceptual discussion ofBowlby's fivetherapeutic tasks is enriched when adult attachment status, as measured by the Adult Attachment Projective, is taken into consideration with the Social Cognitionand Object Relations Scale (Westen, 1995). In relating attachment status to object relations;the attachment classificationcan be translated beyond a construct of the mind, into behavioral factorsinfluenced by the individual's unconscious, internalworld. Even in a young adult college population, differencesin the internalworking models· of secure and insecure individuals can be· considered in a clinical frameworkof seven out of eight object relations dimensions. The only dimension of the SCORS which did not show a significantrelationship between attachment status and object relations, was Emotional Investment•in Values and Moral

Standards. This result was not particularly surprisingbecause the V3!iabledid not seem sensitive enough to detect differencesin a young adult college population. There is also no specificliterature in attachment theory which addresses ethical and moral behavior in relationship to guilt and remorse, or the ability to think abstractly about ethical issues or challenge conventionality and social norms, as being relatively different across the attachment categories.

The surprising aspect of this data set was the large nu111ber of individuals classifiedas insecure. A common concept in attachment theory, supported by empirical evidence, is the expectation that the majority of the population would have secure classificationsthat are stable over time (Levy, Blatt, & Shaver, 1998). Over half the

61 subjects in this sample qualified for some formof insecure classification. This discovery seems to supportF onagy' s postulations that attachment status is more malleable than formerly�elieved as a fixedand categorical construct, and perhaps may be subject to environmental factors and developmental stages at certainages ( 1999).

Meanwhile, the same longitudinal study which provided evidence forthe stability of adult attachment classificationcategories over a twenty year period also showed that 44% of the classificationswhich did change over time were related to negative environmentaJ

]ifefactors, such as parental divorce (Waters et al., 2000). The researchers concluded there was supportfor Bowlby 's position that attachment classifications are flexible and contingent upon environmental forces(Bowlby, 1988). It seemed possible that the attachment system of the subjects in this sample maype overwhelmedor at least in a state . . of fluctuation due to a recent death of someone close, being in college, or having experienced parental divorce. This··study was able to make some preliminary investigations of the relationship between adult attac�ment status and negative environmental factors.

According to the demographicinformation, many students in this population experienced the death of a grandparent, aunt, uncle, or close friendwithin the past three years. It appeared fromthe demographicsheet that this may be the individual's first experience with· loss due to death, andquite possibly very disruptiveto his or her attachment system if they are dealing with these issues for the first time in their life.

Statistical analysis showed that an equal number of students in the secure and insecure categories experienced loss of someone close to them due to death in the past three years,

62 rather than a higher number of insecure individuals experiencing this type of loss, as one

would expect. Perhaps an individual's attachment style of secure or insecure influences

how they contend with loss and creates room forvaried individual differencesnot

captured in this analysis.

Being launched into a more independent setting, such as college, where many

students may be experiencing their firstreal separation fromfamily, may impact or

disrupt the attachment system. Students have lost the security of their immediate family

relations, as well as friendsthrough childhood, and the predictability of a.high school

environment. They are thrust into a new world, where they must conquer the challeng(\5

of a college environment. As students navigate these new interpersonal relationships

with teachers and peers, one could expect that their attachment systems would be

disrupted. Yet again, statistical analysis showed that there were roughly equal numbers

of secure and insecure individuals distributed throughout the freshman, �ophomore, junior, and senior level classes. Although no definitive pattern �ould be detennined

through analysis, differences in attachment style may relate to how a �tudent �opes with

college.

One could theorize that if attachment statusis stable over time, an individual's

attachment status and internalworking model would aid them in contending with life's

losses and transitions, such as death of someone close or entering college. But the

question regarding the large number of insecure classificationsremained unanswered.

According to theory, maladaptive internalworking models exist as a means to defend

against the severe andoverwhelming anxiety thatresults fromloss or separation from a

63 parent. There are fewstudents in this sample who actuallylost a parent due to death or early separation from birth, but many students who received an insecure attachment classification. The effectsof divorce or separa�ion would seem to impact the attachment system, as the parent may no longer be available to the child in a stable and predictable fashion. Quite possibly the parent became unavailable to the child some time prior to the actualdivorce because of the upheaval surrounding personal and marital issues. Again, statistical analysis revealed no meaningfuldistributions between secure and insecure classifications in relationship to parentaldivp rce. There tnftYbe unique components to an individual's attachment system that shape how one deals with parental divorce that were not captured�by this study.

It is important to keep in mind, that these explorations of other meaningfullosses, year in college, and parental divorce are based solely o:µ simple yes or no demographic questions. Previous research exploring other negative environmental litefactors used detailed narratives from the AAI and also tapped more sensitive issues such as physical and sexual abuse (Waters et al., 2000). Overall, the large number of insecure classificationssupports the idea that attachment classificationsare malleable and subject to environmental factors or certain deveJopmental ages. Although a few,simple demographic questions may not capture the fullimp�ct or nuances of these factors, this study brings attention to the need forfuture research in these areas. Other potential lines of researchinclude using other various measures to tap differentvarying levels of conscious andunconscious emotional and behavioral functioning. For example, future research with this data set will include exploring the particular defensive processes of

64 secure and insecure individuals. Meanwhile, self-report inventories on most any dimension of psychological and behavioral processes may illuminate more differences between secure and insecure individuals by taking attachment classifications beyond a theoretical construct.

Limitations of this study include the sample size and ethnic and gender components of thepopulation. A larger sample size would allow "ror more finelytuned comparisons among the insecure classifications of dismissive, preoccupied, and unresolved. This study had fewermales than females and extremely low ethnic diversity.

Although no gender differences were detected, such findingsmight be impacted with a larger sample size of males. Regardless, it is next to impossible to generalize these results to other ethnic populations, and certainly more research is required in this area.

One may consider the small mean differences as a potential weaknessto the study. On the contrary, because the study found significantdiff erences between averages in a young adult college population speaks to the strength of the study, as well as the abiiity of the

Social Cognitionand Object Relations Scale to detect subtle differences i11 a normative, young adult population. An excellent follow-up study would be to compare Adult

Attachment Projective protocols and Thematic Apperception Test protocols from a clinical sampleto this sample. In this way, a more comprehensiveunderstanding of clinical implications could be empirically achieved, while larger mean differences would most likely be obtained.

Unique strengths of this study include contributing to a more comprehensive understandingof human development by investigating similaritiesacross attachment

65 theory, classical psychoanalytic theory, and object relations. The research used an innovative new projective instrument, the Adult AttachmentProjective, with the

Thematic Apperception Test and Drew Westen's Social Cognition and Object Relations

Scale, to empirically link attachment theory with object relations in a young adult normativepop ulation. The findingsof this study expandattachment theory through further definingJohn Bowlby's concept of the internal working model with differences in object relations profiles of secure and insecure individuals, while providing initial evidence that attachmentclassifications are malleable across time.

66 List of References

67 Ackerman, S., Clemence, A., Weatherill, R., & Hilsenroth, M. (1999). Use of the TAT

in assessment of DSM-IV cluster B personality disorders. Journal of Personality

Assessment, 73 (3), 422-448.

American Psychological Association. (2002). Publication Manual of the American

Psychological Association. Washington, DC: American Psychological

Association.

Aviezer, 0., Sagi, A., Joels, T., & Ziv (J.999). Emotional availability and attachment

representations in Kibbutz.infantsand their mothers. Developmental

Psychology. 35 (3), 811-821 .

. Baker_smans, M. & vanIJzendoom, M. ( 1993). A Psychometric study of the Adult

Attachment Interview reliability and discriminant validity. De'{elopmental

Psychology, 29 (5), 870-879.

Bell, M., Billington, R., Cicchetti, D., & Gibbons, J. (1988). Do object re]ations deficits

distinguish BPD fromother diagnostic groups? Journalof Clinical Psychology,

44 (4), 511-5 16.

Bowlby, J. (1969). Attachment and Loss Vol. I: Attachment. New York, New York:

Basic Books, lnc.

Bowlby, J. (1973). Attachment and Loss Vol. II: Separation. New York, New Yc>rk:

Basic· Books, hie.

Bowlby, J. (1980). Attachment and Loss Vol. III: Loss. New York, New York:

Basic Books, Inc.

Bowlby, J. (1988). A Secure Base Parent-Child Attachment and Healthy Human

68 Development. United States of America: Basic Books, Inc.

Cassidy, J. (1998). Attachment andobject relations theories and the concept of

independent behavioral systems. Social Development, 7 (1), 120-126.

Cortina, M. (2001). Sullivan's contributions to understanding personality development

in light of attachment theory and contemporary models of the mind.

ContemporaryPsychoanalysis, 37 (2), 193 -238.

Cowan, P., Cohn, D., Cowan, C., & Pearson, J. (1996). Parents' attachment histories

And children's externalizingand internalizingbeh aviors. Exploring family

systems models of linkage. Journalof Consulting and Clinical Psychology,

64 (1), 53-63.

DeKlyen, M. (1996). Disruptive behavior disorder and intergenerational attachment

Patternsa comparisonof clinic-referredand normally functioningprescho olers

andtheir mothers. Journal of Consulting and Clinical Psychology, 64. (2), 357-

� 365.

DeMulder, E., Denham, S., Schmidt, M., & Mitchell, J. (2000). Q-Set Assessment

Of attachmen� security during the preschool years: Links home to school.

Developmental Psychology, 36 (2), 274-282.

Dfamond, D., Clarkin, J., Levine, H., Levy� K., Foelsch, P., Yeomans, F. (1999).

Borderline conditions and attachment: A preliminary report. Psychoanalytic

lnguiry, 19 (5), 831-884.

Freud, S. (1895). Project fora scientificpsychology. In P. Gay (Ed.), The Freud reader

(pp. 86-89). New York, NY: W.W. Norton & Company,Inc.

69 Freud, S. (1911 ). Formulations on the two principles of mental functioning. In P. Gay

(Ed.), TheFreud reader (pp. 301-309). New York, NY: W.W. Norton &

Company, Inc.

Freud, S. (1920). Beyond the pleasure principle. In P. Gay (Ed.), The Freud reader

(pp. 594-626). New York, NY: W.W. Norton& Company, hlc.

Freud, S. (1923). The ego and the id. In P. Gay (Ed.), The Freud reader (pp. 628-661).

New York, NY: W.W. Norton & Company, Inc.

Freud, S. (1924). Three essays on the theoryof sexuality. In P. Gay (Ed.), The Freud

reader (pp. 239-293). New York, NY: W.W. Norton & Company, Inc.

Fonagy, P. (2001). Attachment theoryand Psychoanalysis. New York, NY: Other

Press.

Fonagy, P. (1999). Psychoanalytic theory from the viewpoint of attachmenttheo ry and

Research. In J. Cassidy & P. Shaver (Eds.), Handbook of attachment: Theory,

research, and clinical applications(pp. 595-624). New York, NY: Guilford

Press.

Fonagy, P., Leigh, T., Steele, M., Steele, H., Kennedy, Il., & Mattoon, G. (1996).

The relation of attachment status, psychiatric classification, and response to

psychotherapy. Journalof Consulting and Clinical Psychology, 64 (1 ), 22-31.

George, C. & Solomon, J. (1996). Representational models of relationships: Links

between caregiving and attachment. InfantMental Health Journal,17 (3), 198-

216.

George, C. & West, M. (2001 ). The development and preliminary validation of a new

70 measure of adult attachment: TheAdult Attachment Projective. Attachment &

HumanDevelopment, 3 ( 1 ), 30-6 1.

George, C., West, M., & Pettem, 0. (1999). The Adult AttachmentProjective.

Unpublished attachment measure and coding manual. Mills CoJlege, Oakland,

CA.

Gilliom, M., Shaw, D., Beck, J., Schonberg9 M., & Lukon, J. (2002). Anger regulation

In disadvantaged preschool boys strategies, andtecdents, and.the development of

Self-control. Developmental Psychology, 3&_(2), 222-235.

Hibbard, S., Hilsenroth, M, Hibbard, J, & Nash, M� (1995). A validity study of two

projective object representations measures. Psychological Assessment, 7 (4),

432-439.

Klein, M. (1952). Some Theoretical Conclusions Regarding the Emotional Lifeof the

Infant. Envy and Gratitude (pp. 61-75). London: HogarthPre ss.

Kohut, H. ( 1977). Does psychoanalysis need a psychology of the selfl fu._ TI1e

Restoration of The Self (pp. 63-139). New York: InternationalUnivers ities

Press, Inc.

Koren, K., Oppenheim, D., Dolev, S·., Sher, E., Etzion-Carasso, A. (2002). Mother's insightfulnessregarding their infants' internalexperience s: Relations

With maternalsensit ivity and infant attachment. Developmental Psvchology, 38

(4), 534-542.

Levy, K., Blatt, S. & Shaver, P. (1988). Attachment styles and parental representations.

Journalof Personalityand Social Psychology, 74 (2), 407-419.

71 Laible, D. & Thompson, R. (1998). Attachment and emotional understanding of

Preschool children. Developmental Psycholo� (5), 1038-1045.

Lyons-Ruth, K. (1996). Attachment relationships among chi1dren ,vith aggressive

behavior problems: The role of disorganized early attachmentpatterns. Journal

of Consulting and Clinical Psychology, 64 (1), 64-73.

Moss, E. & St.-Laurent, D. (2001). Attachment at school age and academic

Performance. Developmental Psychology, 37 (6), 863-874.

Muir, R. (1995). Transpersonal processes: A bridge between object relations and

attachment theory in normal and psychopathological development. British

Journalof Medical Psychology, 68, 243-257.

Murry, H. (1943). Manual forthe Thematic Apperception Test. Cambridge, MA:

HarvardUniversity Press.

Pasoda, G., Jacobs, A., Richmond, M., Carbowell, 0., Alzate, G., Bustamante, M., &

Quicino, J. (2002). Maternal caregiving and infantsecurity in two cultures.

Developmental Psychology, 38 (1), 67-78.

Pianta,R., Egeland, B., & Adam, E. (1996). Adult attachment classificationand self:.

reported psychiatric symptomatologyas assessed by the Minnesota Mulitphasic

PersonalityInventory-2. Journalof Consulting and Clinical Ps�hology. 64

(2), 273-281.

Rothbard, J. & Shaver, P. (1994). In Attachment in Adults: Clinical and Developmental

Issues (pp. 31-71). New York, NY: GuilfordPress.

Sable, P. (1992). Attachment theory: application to clinical practice with adults.

72 Clinical Social Work Joumal,32 (5), 271-283.

Slade, A., Belsky, J., Aber, J., & Phelps, J. (1999). Mothers' representationsof their

Relationships with their toddlers links to adult attachment and observed

mothering. Developmental Psychology, 35 (3), 611-619.

Starns,G, Juffer, F., & van ljzendoorn, M. (2002). MaternalSensitivity, infant

Attachment, and temperament in early childhood predict adjustment in middle

Childhood. The case of adopted children and their biologically unrelated parents.

Developmental Psychology, 38 (5), 806-821.

Stuart, J., Westen, D., Lohr, N., Benjamin, J., Becker, S., Vorus, N., &. Silk, K. (1990).

Object relations in borderlines, depressives, and normals: An examination of

human responses on the Rorschach. Journalof PersonalityAssessment, 55 (1&2),

296-318.

Silverman, D. (1991). Attachment patterns arid Freudian th�ory: An integrative

Proposal. PsychoanalyticPsycholog y, 8 (2), 169-193.

Tramantano, G., Javier, R., & Colon, M. (2003). Discriminating among subgroups of . borderline personality disorder: An assessment of object representations. The

AmericanJournal of Psychoanalysis, 63 (2), 149-175.

Waters, E., Merrick, S., Treboux, D., Crowell, J. & Albersheim, L. (2000). Attachment

securityin infancy andearly adulthood: A twenty-year longitudinal study.

Child Development,71 (3), 684-689.

West, M. & George, C. (2002, in press) Attachment and dysthymia: The contributions

73 of preoccupied attachment and agency of self to depression in women. Attachment and

Human Development.

Westen, D. (1991). Clinical assessment of object relations using the TAT. Journalof

PersonalityAssessm ent, 56 (1 ), 56-74.

Westen, D. (1995). Social Cognition� d Object Relations Scale: O-Sort forProjective

Stories. Cambridge Hospital andHarvard Medical School: Cambridge, MA.

Westen, D., Lohr, N., Silk, K., Gold, L., & Kerber, K. (1990). Object relations and

social cognitionin borderlines, depressives, and normals: A Thematic

Apperception test analysis. Psychological Assessment, 2 (4), 355-364.

Winnicott, D. ( 1931). A note on normality and anxiety. In Through paediatrics

to psychoanalysis collected papers (pp. 3-22). London: Kamac Books.

Winnicott, D. (1935). The manic defence. In Through paediatrics to psychoanalysis

collected papers (pp. 129-145). London: KarnacBooks.

W"innicott, D. (1941). The observation of infantsin a set situation. In Through

paediatrics to psychoanalysis col1ected papers (pp. 52-70). London: Kamac

Books.

Winnicott, D. (1948). Reparation in respect of mother's organized defenceagainst

depression. In ThroughPaediatrics to psychoanalysis collected papers (pp. 91-

97). London: Kamac Books.

Winnicott, D. ( 1952). Psychoses and child care. ln Throughpaediatrics to

psychoanalysis collected papers (pp. 219-229). London: Kamac Books.

Winnicott, D. (1960). The theory of the paren�-infantrel ationship. In The

74 maturational processes and the facilitatingenvironment studies in the theory,

of emotional development (pp. 37-56). Madison, CN: InternationalUniversities

Press, Inc.

Winnicott, D. (1960). Ego distortion in terms of true and false self. In The maturational

processes and the facilitating environmentstudies in the theoryof emotional

development (pp. 140-153). Madison, CN: International Universities Press, Inc.

Winnicott, D. (1962). A personal view of the Kleinian contribution. In lbe

Maturational processes and the facilitatingenvironment studies in the theory

Of emotional development (pp. 171-179).

Winnicott,D. (1971). Playing and reality. New York, NY: Routledge.

75 Appendix A

76 Table 1 Kappa and Intra-class Correlation Coefficients_

For the Interrater Reliability of the AAP and SCORS

Measures Reliability Rate p-value Adult Attachment Projective .87 .0000 Multi-rater Kappa, 5 raters

Thematic Apperception Test Intra-class Correlation Coeffi.cients, 4 raters

Complexity of Representations of People .87 .0000

AffectQuality of Representations .94 .0000

Emotional InvestmentIn Relationships .96 .0000

Emotional Investmentin Values and Moral .81 .0000 Standards

Understanding of Social Cam,ality .84 .0000

Experience and Management of Aggressive .92 .0000 Impulses

Self-Esteem .75 .0000

Identity and Coherence of Self .71 .0002

77 Table 2 Levene's Test 9f Equality of Error Variances

F df Significance

Complexity of Representation of People .014 1 .907

Affective Quality of Representations .398 I .529

Emotional Investment in Relationships .313 1 .577

Emotional Investment in Values and Moral .037 1 .847 Standards

Understanding of Social Causality .360 1 .550

Experience and Management of Aggressive 3.730 1 .056 Impulses

Self-Esteem 1.920 1 .169

Identity and Coherence of Self 3.023 .085

78 Table 3 Sample Demographic Data

Frequency Percent Cumulative Percent Gender Male 28 28.9 28.9 Female 69 71.1 100.0 Total 97 100.0 Ethnic Background Caucasian 87 89.7 89.7 AfricanAmerican 7 7.2 96.9 Asian 3 3.1 100.0 Total 97 100.0 Year in College Freshman 32 33.0 33.0 Sophomore 11 11.3 44.3 Junior 25 25.8 70.1 Senior or Above 29 . 29:9 100.0 Total 97 100.0 Raised by Parents Yes 91 93.8 93.8 No 6 6.2 100.0. Total 97 100.0 Loss of Parent Due to Death or Separation Yes 4 4.1 4.1 No 93 95.9 100.0 Total 97 100.0 Experience of Parental Divorce Yes 27 27.8 27.8 No 70 72.2 100.0 Total 97 100.0 Loss of Another Relative Due to Death Yes 46 47.4 47.4 No 51 52.6 100.0 Total 97 100.0 Experience of Major Personal Illness or Injury Yes 7 7.2 7.2 No 90 92.8 100.0 Total 97 100.0

79 Table 4 Pearson's R Correlations of Demographic Data and SCORS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1 1.00 .007 ....224* .298** .290**. -.037 .109 .164 .177 .123 .150 .138 .083 .159 .110 .945 .027 .003 .004 .72] .289 .109 .083 .232 .141 .177 .419 .119 .284 2 --- 1.00 .120 -.097 -.061 .002 .195 .074 · -.013 J60 -.056 .074 -.108 -.158 -.037 .243 .346 .551 .986 .056 .468 .896 .118 .585 .469 .294 .122 .72 1 3 ------1.00 -.592** -.222* -.094 -.013 -.069 -.117 -.0] 1 -.150 -.022 -. 153 -.172 -. 137 .000 .029 361 .897 .502 . .252 .916 .143 .828 .136 .092 .180 00 4 ------1.00 .103 -.058 .218* .023 -.055 -.062 .131 -.008 -.020 .116 .105 .317 .574 �U32 .821 .591 .545 .202 .940 .847 .257 .307 5 --- .. __ ------1.00 .005 -.037 .087 -.019 -.100 -.026 .060 .008 -.002 -.041 .964 :n9 .398 .852 .332 .803 .561 .935 .987 .693 6 ------1.00 -.105 .018 .102 .055 .103 .064 .091 .039 -.015 .305 .86 1 .320 .590 .315 .535 .378 .707 .883 7 ------1.00 .021 -.083 -. 101 .094 -.076 -.176 .022 .095 .837 .417 .325 .361 .459 .084 .832 .354

Note. *p<.05, **p<.01 N=97 foreach ceH. 1 = Age, 2 = Gender, 3 = Raised by Parents, 4 = Loss of Parent, 5 = Parental Divorce, 6 = Personal Injuryor Illness, 7 = Death of Someone Close, 8 = Complexity of Representationof People, 9 = Affective Quality of Representations, 10 = EmotionaJ Investment in Relationships, 11 = Emotional Investment in Values and Moral Standards, 12 = Understanding of Social Causality, 13 = Experience and Management of Aggressive Impulses, 14 = Self-Esteem, 15 = Identity and Coherence of Self. Table 5 AAP & Other Death Cross Tabulations and Chi-Square Test

Adult Attachment Other Other Loss- Total Pearson Chi df Asymp. Sig. Loss-Yes No Square Value (2-sided) Secure Count 23 21 44 .760 1 .383 % Within AAP 52.3 47.7 100.0 % Within Other Death 50.0 41.2 45.4 % of Total 23.7 21.6 45.4

Insecure Count 23 30 53 % Within AAP 43 .4 56.6 100.0 % Within Other Death 50.0 58.8 54.6 -00 % of Total 23.7 30.9 54.6 Total Count 46 51 97 % Within AAP 47.4 52.6 100.0 % Within Other Death 100.0 100.0 100.0 % of Total 47.4 52.6 100.0 Table 6 AAP & College Year Cross Tabulatfons and Chi-Square Test

Adult Attachment Freshman Sophomore Junior Senior or Total Pearson Chi df Asymp. Sig. Above Square Value (2-sided) Secure Count 13 6 13 12 44 1.294 3 .73 1 % Within AAP 29.5 13.6 29.5 27.3 100.0 % Within Other Death 40.6 54.5 52.0 41.4 45.4 % of Total 13.4 6.2 13.4 12.4 45.4

·. Insecure Count 19 5 12 17 53 00 % Within AAP 35.8 9.4 22.6 32.1 100.0 % Within Other Death 59.4 45.5 48.Q 58.6 54.6 % of Total 19.6 5.2 12.4 17.5 54.6

Total Count 32 11 25 29 97 % Within AAP 33.0 11.3 25.8 29.9 100.0 % Within Other Dea.th 100.0 100.0 100.0 100.0 100.0 % of Total 33:0 11.3 25.8 29.9 100.0 '. I

Table 7 AAP & Parental Divorce Cross Tabulations and Chi-Square Test

Adult Attachment Parental Parental Total Pearson Chi df Asymp. Sig. Divorce- Divorce- Square Value (2-sided) Yes No Secure Count 10 37 44 1.046 1 · .306 % Within AAP 22.7 77.3 100.0 % Within Divorce 37.0 48.6 45.4 % ofTotal 10.3 35.1 45.4

Insecure Count 17 36 53 % WithinAAP 32.1 67.9 100.0 00 % Within Divorce 63 .0 51.4 54.6 % of Total 17.5 37.1 54.6 ' Total Count 27 70 97 % Within AAP 27.8 72.2 100.0 % Within Divorce 100.0 100.0 100.0 % of Total 27.8 72.2 100.0

., Table 8 Descriptive Statistics and Estimates of AAP and SCORS •

Dependent Variable Adult M STD N 95% ConfidenceInterval Attachment Lower Upper

Complexity of Secure 4.291 .704 44 4.073 4.508 Representation of Insecure 3.879 .745 53 3.681 4.077 People Total 4.066 .752 97

Affective Quality of Secure 4.000 .578 44 3.829 4.171 Representations Insecure 3.479 .565 53 3.324 3.635 Total 3.716 .625 97

Emotional Secure 3.805 .717 44 3.593 4.016 Investment in Insecure 3.374 .696 53 3.181 3.566 Relationships Total 3.569 .735 97

Emotional Secure 3.986 .399 44 3.880 4.093 Investment in Values Insecure 3.875 .314 53 3.779 3.972 and Moral Standards Total 3.926 .357 97

Understanding of Secure 4.200 .820 44 3.968 4.432 Social Causality Insecure 3.706 .739 53 3.494 3.917 Total 3.930 .812 97

Experience and Secure 3.964 .396 44 3.833 4.094 Management of Insecure 3.653 .468 53 3.534 3.772 Aggressive Impulses Total 3.794 .461 97

Self-Esteem Secure 4.141 .480 44 4.009 4.273 Insecure 3.872 .'408 53 3.751 3.992 Total 3.994 .460 97

Identity and Secure 4.882 .432 44 4.719 5.044 Coherence of Self Insecure 4.476 .621 53 4.327 4.624 Total 4.660 .578 97

84 Table 9 Comparisons of Secure and lose.cure Findings of the AAPand SCORS

Dependent Variable Mean Sig. 95% ConfidenceInterval Difference Lower Bound Upper Bound Secure- Insecure Complexity of .412* .007 .118 .706 Representations of People

AffectiveQuality of .521 * .000 .290 - .752 Representations

Emotional Investment in .43 1 * .004 .145 .717 Relationships

Emotional Investment in .111 .129 .033 .255 Values andMoral Standards

Understanding of Social .494* .602 ·.180 .809 Causality

Experienceand .311* .001 .134 .488 Management of Aggressive Impulses

Self-Esteem .269* .004 .448

Identity and Coherence .406* .000 .186 .626 of Self

Note. *p<.01

85 Vita

th Betty Marie Martin was bornin Okinawa, Japan on Ap1il 8 , 1973. At the age of two, she moved to the United States with her father. Bettygrew up in Lebanon,

Pennsylvania and always had an affinity forworking with people. While still attending

Cedar Crest High School, she became employed as a certified nurse assistant to. elderly individuals in a hospice. She graduated in 1991 and completed a dual major of psychology and philosophy at Millersville University in Millersville, Pennsylvania. In college, she was employed as the Adoption Project Research Coordinator, volunteered at an outpatient, partial hospitalization program forindi viduals with severe mental illness, and became a residential rehabilitation specialist foradults with persistent and chronic menta_l illness in a group home. She graduated cum laude fromMill ersville in 1995.

Betty later pursued her doctoral degree in Clinical Psychology at the University of

Tennessee, Knoxville. She completed research in the areas of projective testing, attachment theory, and object relations. As part of her training, she worked as a therapist to low-income populations through the university clinic and a local community mental health center. Betty was also a graduate teaching instructor and taught eightseme5:ters of psychology courses including honors introductory psychology, general mtroductory psychology, abnormal psychology, and child psychology. She was awarded a Mark of

Distinction for Honors Quality in Graduate Clinical Practice and the University of

Tennessee Clinical Psychology Alumni A ward forExcellence in Scholarship. She successfullydefended her dissertation in June of 2004 and will be completing her internship at the State University of New York in Stony Brook, New York.

86 45350545p 25('J 11/19/BS t t113'